Abstract PD3-07: A phase II pre-surgical trial of lapatinib for the treatment of women with HER2 positive or EGFR positive ductal carcinoma in situ

Background: Estrogen receptor (ER)-negative tumors and human epidermal growth factor 2-Neu (HER2) positive breast cancers are known to be more clinically aggressive subtypes of breast cancer and account for 30% of all breast cancers. Women with HER2 + breast cancers, whether ER+ or ER -, require cytotoxic chemotherapy with a HER2-targeting agent, and often have adverse outcomes. Thus, preventive agents are needed to reduce the incidence of these subtypes of aggressive breast cancer. Lapatinib, a dual tyrosine kinase inhibitor, inhibits epidermal growth factor receptors (EGFR) and HER2 kinases and has shown to decrease breast cell proliferation in invasive breast cancer and adjacent premalignant lesions. Therefore, we conducted a multi-institutional randomized Phase II clinical trial to study the effects of the signal transduction inhibitor lapatinib in women with HER2-positive or EGFR-positive ductal carcinoma in situ (DCIS).

Methods: Randomized participants received either lapatinib (750mg, 1000mg, or 1500mg) or placebo daily for 2-6 weeks prior to their surgery. After minimal accrual, the trial was later amended to lapatinib 1000mg or placebo. Pre-treatment breast tissue was obtained from initial diagnostic core biopsy and post-treatment breast tissue was obtained from surgical excision specimen. Blood was obtained prior to surgery to assess serum lapatinib level. Participants kept a daily symptom assessment log and had a cardiac assessment at baseline and prior to surgery. Patients were instructed to take drug up to and including the day before surgery. The dual primary endpoint for this study was change in proliferation in pre- versus post-treatment biopsies between the two treatment arms, as measured by Ki67 as well as toxicity assessment. Secondary endpoints included incidence of DCIS at surgery and modulation of tissue biomarker expression in growth factor receptors (EGFR, ErbB2); phosphorylated growth factor receptor (phospho-ErbB2); signal transduction markers (MAPK, phospho-MAPK); hormone receptors (ER, PR); and p27.

Results:Twenty-two women (mean age: 51; range: 32-66) with HER2+ or EGFR+ DCIS were treated with lapatinib (1,000 or 1,500 mg) or placebo for 2–6 weeks prior to surgical excision. Ki67 expression was significantly decreased in the lapatinib treatment arms compared to placebo (p=0.0122). Diarrhea, fatigue, and skin reactions were notable adverse events that occurred predominately in the lapatinib arm compared to placebo. No grade 3 or 4 events related to the study drug were noted during the study. No changes were noted in cardiac function. DCIS was present in all surgical specimens in both arms. Invasive breast cancer was noted in 1 patient on lapatinib 1000mg and 3 patients on placebo. No statistically significant changes were noted in signal transduction biomarkers

Conclusion:These results demonstrate the effectiveness of lapatinib in reducing proliferation in women with EGFR+ or HER2+ DCIS. Even low-grade toxicities can deter use of an agent in the prevention setting. This and the lack of a risk model for HER2+ and triple negative breast cancer make the development of larger scale clinical prevention trials of lapatinib for the prevention a challenge.

Citation Format: Thomas PS, Contreras A, Pruthi S, Krontiras H, Rimawi M, Garber J, Wang T, Hilsenbeck SG, Vornik LA, Gilmer T, Friedman R, Heckman-Stoddard BM, Dunn B, Kuerer H, Brown PH. A phase II pre-surgical trial of lapatinib for the treatment of women with HER2 positive or EGFR positive ductal carcinoma in situ [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr PD3-07.

Abstract P3-01-05: Circulating tumor cell detection predicts early recurrence in patients with non-metastatic breast cancer

Background:

Our group and others have shown that CTCs are identified in 20-25% of non-metastatic breast cancer patients. Recent data suggests that circulating tumor cell (CTC) detection at 5 years follow-up predicts late recurrence for non-metastatic, estrogen receptor positive, HER2/neu negative (ER+/HER2-) breast cancer patients. The aim of this study was to determine if CTC detection at time of surgery in ER+/HER2- predicts early (less than or equal to 5 years) recurrence, and to compare the median recurrence-free survival (RFS) to triple negative (TN) patients.

Methods:

We performed CTC enumeration on 506 patients with non-metastatic breast cancer just prior to surgical resection as part of an IRB approved study. CTCs (per 7.5 ml blood) were identified using the Cell SearchSystem (Menarini Silicon Biosystems). The presence of one or more CTC meeting morphological criteria for malignancy was considered a positive result. Patients with inflammatory breast cancer were excluded from our analysis. Log-rank test and Cox regression analysis were applied to establish the association of CTCs with RFS.

Results:

Median follow-up was 68 months and mean age was 53 years. Eighty-four percent of patients (417/498) had T1/T2 tumors, 63% (307/487) had grade 1 or 2 tumors, and 58% (292/501) had negative lymph nodes. Eighty-three percent (419/506) had ER+/HER2- tumors, and 17% (87/506) were TN. One or more CTC was identified in 22% (91/419) ER+/HER2- patients, and in 28% (24/87) of TN patients. Nine percent (45/506) of patients recurred within 5 years of CTC assessment. Of the 45 patients who relapsed, detection of one or more CTCs predicted shortened RFS. Median RFS for CTC-positive patients was 1.2 years, vs. 2.5 years for CTC-negative patients, irrespective of ER+/HER2- or TN subtype (log-rank P<0.001, HR = 2.71, 95% CI, 1.50 to 4.87).

Conclusions: One or more CTCs identified at surgical resection predicted shortened RFS, irrespective of subtype, in non-metastatic breast cancer patients. This data warrants larger studies to determine if CTC positivity can be used to stratify both ER+ and TN patients who are at increased risk for early recurrence.

Citation Format: Hall C, Meas S, Bowman Bauldry J, Kuerer H, Lucci A. Circulating tumor cell detection predicts early recurrence in patients with non-metastatic breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P3-01-05.

Abstract P5-13-03: Factors associated with prophylactic mastectomy among BRCA-positive patients with no personal history of breast cancer

Background: The rate of prophylactic mastectomy (PM) has recently increased1. A deleterious mutation in the BRCA1 or BRCA 2 genes is among the major reasons why patients pursue PM 2, 3. Women with BRCA1 or BRCA2 gene mutations have up to an 87% risk to develop an invasive breast cancer (BC), and up to 45% risk for ovarian cancer (OC)2, 3. Most studies evaluating predictors of PM in BRCA mutation carriers are performed among women with breast cancer; however, accurate predictors for PM among unaffected BRCA mutation carriers are less defined. In a single institution study we aimed to evaluate predictors of PM among BRCA carriers with no personal history of breast cancer.

Method: One hundred seventy seven women with no personal history of BC, who tested positive for a BRCA1 or BRCA 2 germline mutation, were included in the study. Patients’ characteristics were obtained from a prospectively maintained research database under an IRB approved protocol at UT MD Anderson Cancer Center. Univariate analyses using chi-square and logistic regression analysis were used to determine predictive factors associated with PM. The patient characteristics examined included age, martial and educational status, bilateral salphingo-oophorectomy (BSO), family history of 1st and 2nd degree relatives with breast (BC) and ovarian cancer (OC), race, and BRCA genetic test result.

Results: Out of the 177 BRCA1 and BRCA 2 positive patients, 51 (29%) elected for PM. The average age for the cohort was 44 years (range 23-91). The majority were Caucasian (81%), and married (72%) with a college degree (64%). One hundred sixty-four (92%) patients had 1st and 2nd degree relatives with BC, 93 (53%) had 1st and 2nd degree relatives with OC, and 85 (48%) had undergone BSO. A logistic regression model was run to identify factors associated with undergoing PM, including family history of OC, family history of BC, BSO and age. Patients with a family history of OC were 2.5 times more likely than those without to have had a PM (p = 0.0125), and patients who had a BSO were 0.137 times less likely to have had a PM (p <.0001). Only 12 patients did not have a family history of BC and none of those patients had a PM, so an Odds Ratio could not be calculated. However, it was determined through the Fisher's exact test that patients with a family history of BC were more likely to undergo a PM (p = 0.0198).

Conclusion: The rate of PM in our cohort was slightly lower than expected. Factors associated with PM included a family history of BC and OC. Interestingly, having had a BSO was associated with lower likelihood of undergoing PM (menopausal status will be further evaluated); possibly due to the knowledge that BSO reduces breast cancer risk.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P5-13-03.

P4-10-06: Evaluation of BRCAPro Risk Assessment Model in Patients with Ductal Carcinoma In Situ

Background: The BRCAPRo model is used to predict a patient's likelihood to possess a BRCA1 or BRCA2 gene mutation based upon personal and family history. Ductal carcinoma in situ (DCIS) is considered a non-invasive condition which can progress to an invasive breast cancer if left untreated. Currently, DCIS is not specifically accounted for in the BRCAPro model, thereby causing DCIS to be weighted in the same manner as an invasive breast cancer diagnosis. Historically, a diagnosis of DCIS has been entered as having developed into an invasive breast cancer ten years later. However, there are no standard guidelines of how DCIS should be entered. We sought to determine if there were any differences in how DCIS was treated in the BRCAPro model to predict the more effective method in calculating the BRCAPro. Methods: Women with pure DCIS, who were referred for genetic counseling and underwent genetic testing, were included in the study. The likelihood of carrying a BRCA mutation was calculated using the BRCAPRO model (Version 5.1). Patient characteristics which were entered into the BRCAPro model include: tumor markers (estrogen receptor-ER and progesterone receptor-PR), history of oophorectomy prior to diagnosis, family history of 1st and 2nd degree relatives with breast and ovarian cancer, race and Ashkenazi Jewish ancestry. Each patient's BRCAPro risk estimate was calculated and compared by entering DCIS at the presenting age of diagnosis and by adding 10 years to the age of diagnosis. Descriptive statistics and a student's t-test were used to compare BRCAPro estimates between the two groups. Results: Ninety-five patients with pure DCIS underwent genetic counseling and testing. The average age of DCIS diagnosis was 45 years (range 26–65). Of the 95 DCIS patients included in the study 21% (n=20) tested positive for a BRCA gene mutation (8 BRCA1 and 12 BRCA2), 77% (n=74) test negative and 0.01% (n=1) had a variant of uncertain significance. Overall, DCIS patients who tested positive for a BRCA mutation had a higher BRCAPro (40%) than patients who tested negative (12%) when presenting age of diagnosis was assessed. When 10 years was added, the BRCAPro estimate was still higher amongst BRCA positive patients (28%) than BRCA negative patients (8%). The mean BRCAPro probability when DCIS was entered at presenting age of diagnosis was 18% (0.1−95.4) versus 12% (0.1−89.9) when calculated 10 years later. Conclusion: In our cohort there was no significant difference in BRCAPro probability whether DCIS was entered at the presenting age of diagnosis or 10 years later (p=0.1). However, future studies are needed to determine the most effective method to incorporate DCIS into the BRCAPro model in order to determine those individuals who may or may not be at increase risk to possess a BRCA gene mutation.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-10-06.

P4-07-11: Circulating Tumor Cells after Neoadjuvant Therapy Predict Outcome in Stage I to III Breast Cancer

Introduction: Circulating tumor cells (CTCs) predict outcome in metastatic breast cancer, but their significance is unclear in non-metastatic patients. Furthermore, it is unclear if the presence of CTCs after completion of neoadjuvant chemotherapy (NACT) predicts worse outcome. The purpose of this study was to determine if CTCs after NACT predicts worse outcome.

Methods: Clinical stage I-III breast cancer patients seen at a single tertiary cancer center provided informed consent to participate in an IRB-approved study involving collection of blood (7.5 ml x 2 tubes) at the time of surgery for their primary breast cancer. CTCs were detected using the Cell SearchTM system. A positive result was defined as the presence of one or more cells per 7.5 ml blood since the threshold for positivity has not been established in non-metastatic breast cancer. Statistical analyses used chi-square and Fischer's exact test.

Results: One hundred and twenty patients were prospectively enrolled. Median age was 50 years and median follow-up was 33 months. Eight percent of patients had T1 disease, 35% T2, 18% T3, and 39% T4. Fifty-three percent of patients (63/120) had hormone receptor positive disease. Thirty-two percent of patients (38/120) were HER-2 positive. Thirty percent (36/120) were triple negative. Seventy-eight percent (91/120) had lymph node positive disease. Two or more CTCs were present in 9% of patients (11/120). Of the 11 patients who died, 3 had 2 or more CTCs (P=0.08). Of the 20 who relapsed, 6 had 2 or more CTCs (P=0.0019).

Conclusions: Presence of two or more CTCs after NACT predicted worse relapse free survival in patients with stage I-III breast cancer.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-07-11.

P4-07-07: Circulating Tumor Cells Predict Survival in Non-Metastatic Breast Cancer

Background: Circulating tumor cells (CTCs) predict outcome in metastatic breast cancer (BC), but their prognostic significance in non-metastatic BC is unclear. This study determined whether CTCs could be identified in non-metastatic BC patients and 2) if CTCs predict relapse-free and overall survival.

Methods: Clinical stage I-III BC patients seen at a single tertiary cancer center provided informed consent to participate in an IRB-approved study involving collection of blood (7.5 ml x 2 tubes) before systemic therapy and at the time of surgery for their primary BC. CTCs were detected using the Cell SearchTM system. A positive result was defined as the presence of two or more cells per 7.5 ml blood since a threshold for positivity has not been established in primary BC. Statistical analyses were done using STATA-IC 11 software.

Results: We prospectively evaluated 291 patients. Mean age was 54 years, and median follow-up was 30 months. 54% (157) had T1 tumors, 36%(104) T2, 6% (18) T3, and 4%(12) had T4 disease; 37% (107/289) were clinically node-positive (LNs) by axillary ultrasound and FNA. Two or more CTCs were identified in 10% (29) patients. Seventy-seven percent (225) patients were hormone receptor positive, 11% (32) were HER2 positive while 16% (48) expressed no receptors. Sixteen percent (48) were tumor grade 1, 50% (143) were grade 2 and 34% (98) were grade 3. Systemic chemotherapy and/or endocrine therapy were administered in 80% (233/289) of patients. Sixty-eight percent (167/244) of patients were post-menopausal. Additionally, there was no significant correlation between CTCs with primary tumor size, lymph node status, estrogen or progesterone receptor status, HER2 amplification or tumor grade. Thirty-eight percent (6/16) of all relapses occurred in patients with 2 or more CTCs (HR: 4.48 (95% C.I. 1.61−12.45), Logrank P 0.002) while 40% (4/10) of all deaths occurred in patients with 2 or more CTCs (HR: 4.54 (95% C.I. 1.27−16.25), Logrank P 0.011).

Conclusions: CTCs were a significant predictor of disease-free and overall survival in non-metastatic breast cancer.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-07-07.

P5-01-17: HER2 Amplification in Primary Tumor: A Potential Marker for Presence of Circulating Tumor Cells in Inflammatory Breast Cancer Patients?

Background: Inflammatory breast cancer (IBC) is a rare but aggressive form of invasive breast cancer accounting for 3–6% of all cases and have higher rates of distant recurrence. Circulating tumor cells (CTCs) are known to predict outcome in metastatic breast cancer (BC) patients, but little is known about their prognostic significance in non-metastatic BC. We hypothesized that CTCs can be identified in patients with IBCs and may correlate with primary tumor characteristics. Methods: All patients had blood samples collected at the time of primary surgery. CTCs (per 7.5 ml blood) were detected using the Cell Search™ system (Veridex) and were defined as nucleated cells lacking CD45 but expressing cytokeratins (CK) 8, 18, or 19. The presence of ≥ 1 epithelial cells meeting morphologic criteria for malignancy was considered a positive result. Statistical analyses employed Chi square and Fisher's exact tests using STATA IC 11. Results: We prospectively evaluated 41 IBC patients enrolled in an IRB approved protocol undergoing surgery for stage I-III breast cancer. Median follow-up was 30 months. Mean age was 52 years. Thirty five patients (94%) had positive lymph nodes (LNs) at presentation, 30 (75%) had high-grade tumors and 20 (53%) had lymphovascular invasion. Eleven patients (28%) were ER positive, 11 (27%) were PR positive and 18 (44%) were HER2 positive. IBCs were more likely to be high grade (P<0.0001), ER negative (P<0.0001), PR negative (P<0.0001), HER2 positive (P<0.0001), High Ki-67 (P= 0.005) and had a BMI of more than 25kg/m2 (P=0.04). Eleven (27%) patients were CTC positive. CTCs were more likely be found in HER2 positive (8/18; 44%) vs. HER2 negative primary tumors (3/20; 15%) [OR= 4.53; 95% C.I. = 1.02−19.52; P= 0.04]. We found no statistically significant correlations between primary tumor characteristics (ER, PR, LNs, high grade) and presence of CTCs. Conclusions: About a quarter of IBC patients had CTCs at the time of primary surgery. In these patients HER2 overexpression predicted the presence of CTCs. Studies with longer follow-ups is needed to determine if CTCs predicted survival.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P5-01-17.

P4-06-02: Microscopic Disease in Blood and Bone Marrow Predicts Survival in Early Stage Breast Cancer

Background

Disseminated tumor cells (DTCs) in the bone marrow have been identified in 30% of stage I - III breast cancer (BC) patients and predict survival. Circulating tumor cells (CTCs) in the blood predict outcome in metastatic BC, but their prognostic significance in primary BC is unclear. This study determined whether: 1) DTCs and CTCs could be identified in significant numbers of non-metastatic BC patients and 2) if these cells predict relapse-free (RFS) and overall survival (OS).

Methods: Clinical stage I-III BC patients seen at a single tertiary cancer center provided informed consent to participate in an IRB-approved study involving collection of blood (7.5 ml x 2 tubes) and bone marrow (10 ml from bilateral iliac crests), and at the time of surgery for their primary BC. DTCs were assessed using an anti-cytokeratin antibody cocktail (MNF 116, CK 8,18, and 19, CAM 5.2, and AE1/AE3) following ficoll enrichment and cytospin. A positive result for DTCs was defined by presence of one or more CK-positive cells meeting morphologic criteria for malignancy. CTCs were detected using the Cell SearchTM system. A positive result was defined as the presence of one or more cells per 7.5 ml blood since a threshold for positivity has not been established in non-metastatic BC. Statistical analyses used chi-square and Fischer's exact test.

Results: We prospectively evaluated 313 patients. Mean age was 53 years, and median follow-up was 32 months. Forty-two percent of patients (131) had T1 tumors, 36% (112) T2, 10% (30) T3, and 13% (40) had T4 disease. Forty-five percent of patients (141/312) had positive lymph nodes. DTCs were identified in 29% (91/313) and CTCs in 25% (79/313) of all patients. Seven percent (21/313) of patients had both DTCs and CTCs. In the overall cohort, 26 (8%) patients relapsed and 15 (5%) died. Ten percent (9/91) of DTC positive patients died compared to 3% (6/222) of those who did not have DTCs (p=0.01). Similarly, 6% (7/79) of those who had CTCs died compared to 3% (8/234) of those who did not (p=0.03). Fifteen percent (12/79) of CTC positive patents relapsed compared to 6% (14/234, P=0.01) of those who were CTC negative. Simultaneous presence of DTCs and CTCs was a strong predictor of RFS (log rank p=0.030, HR= 2.8, 95% C.I. 1.20- 8.10) as well as OS (log rank p=0.026, HR= 3.66, 95% C.I. 1.03- 13.00) at 2 years. Combined presence of DTCs and CTCs was a predictor of outcome and these findings persisted after adjusting for variables including hormone receptor status, HER2 status, primary tumor size, grade, and preoperative lymph node status. There was no significant correlation between DTCs and/or CTCs with other primary tumor characteristics.

Conclusions: Circulating and disseminated tumor cells can be identified in a significant number of non-metastatic breast cancer patients. Both CTCs and DTCs predicted outcome, and their combined presence was an independent predictor of survival.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-06-02.

Six-year analysis of treatment-related toxicities in patients treated with accelerated partial breast irradiation on the American Society of Breast Surgeons MammoSite Breast Brachytherapy registry trial

BACKGROUND

The American Society of Breast Surgeons (ASBrS) enrolled women in a registry trial to prospectively study patients treated with the MammoSite RTS device. This report presents 6-year data on treatment-related toxicities from the trial.

METHODS

A total of 1449 primary early-stage breast cancers were treated with accelerated partial breast irradiation (APBI) using the MammoSite device (34 Gy in 10 fractions) in 1440 women. Of these, 1255 case (87%) had invasive breast cancer (IBC) (median size = 10 mm) and 194 cases (13%) had ductal carcinoma in situ (DCIS) (median size = 8 mm). Median follow-up was 59 months. Fisher exact test was performed to correlate categorical covariates with toxicity.

RESULTS

Breast seromas were reported in 28% of cases (35.5% with open cavity and 21.7% with closed cavity placement). Also, 13% of all treated breasts developed symptomatic seromas, and 77% of these seromas developed during the 1st year after treatment. There were 172 cases (11.9%) that required drainage to correct. Use of chemotherapy and balloon fill >50 cc were associated with the development of symptomatic seromas. Also, 2.3% of patients developed fat necrosis (FN). The incidence of FN during years 1 and 2 were 0.9% and 0.8%, respectively. Seroma formation, use of hormonal therapy, breast infection, and A/B cup size were associated with fat necrosis. There were 138 infections (9.5%) recorded; 98% occurred during the 1st year after treatment. Chemotherapy and seroma formation were associated with the development of infections.

CONCLUSIONS

Treatment-related toxicities 6 years after treatment with APBI using the MammoSite device are similar to those reported with other forms of APBI with similar follow-up.

Predictors of Persistent Micrometastatic Disease after Neoadjuvant Chemotherapy

Introduction: Patients who receive neoadjuvant chemotherapy (NAC) for breast cancer typically do not receive further cytotoxic chemotherapy after surgery. We hypothesized that some patients would have micrometastatic disease in the bone marrow (disseminated tumor cells, DTCs) or peripheral blood (circulating tumor cells, CTCs) after NAC. This study documented rates and factors predicting DTCs and CTCs after NAC.Methods: We prospectively evaluated patients undergoing surgery for stage I-III breast cancer. All patients had blood and bone marrow samples taken after completing systemic NAC. CTCs (per 7.5ml blood) were detected using the CellSearchTM system (Veridex). CTCs were defined as nucleated cells lacking CD45 but expressing cytokeratins (CK) 8, 18, or 19. DTCs were assessed using anti-CK antibody cocktail (AE1/AE3, CAM5.2, MNF116, CK8 and 18) following cytospin. The presence of ≥1 CK positive cells and ≥1 epithelial cells meeting morphologic criteria for malignancy was considered a positive result for DTCs and CTCs, respectively. Clinicopathologic factors correlated with DTCs and CTCs and response after NAC included: Her2/neu, estrogen receptor (ER), progesterone receptor (PR) and COX2 status as well as tumor size and grade. Complete pathologic response (pCR) was defined as lack of any residual invasive disease in primary tumor and regional lymph nodes after NAC. Statistical analyses used chi-square and Fischer's exact test.Results: Results were available for 53 patients who had bone marrow and blood collected after NAC. Median follow-up was 26 months. Mean age was 52 years. NAC included anthracyclines and taxanes +/- trastuzumab. Forty percent of patients had either DTCs or CTCs after NAC. Six patients had DTCs amongst 11 patients (55%) who received trastuzumab as compared to those who did not 6/33 (18%), P=0.019. DTCs and CTCs were found in 10/43 patients (23%) and 11/43 patients (27%), respectively. Factors predicting the presence of DTCs after NAC were Her2/neu positivity (P=0.001) and COX2 positivity determined in the primary tumor at diagnosis (P=0.04). No statistically significant correlations were found between CTCs after NAC and primary tumor characteristics. Thirty percent of patients with evidence of DTCs and 10% with CTCs had a pCR. Among the 7 patients with pCR after NAC, 2 (28%) had DTCs and 2 (28%) had CTCs. Factors predicting pCR following NAC were Her2/neu positivity (P=0.0003) and ER negativity (P=0.044). In our analysis, the best predictor of the presence of DTCs and the most likely reason for pCR following NAC was Her2/neu positivity.Conclusions: A significant number of patients have persistent DTCs and/or CTCs after NAC. Interestingly, HER2 positive patients were more likely to have pCR but were also more likely to show persistence of DTCs following NAC. Follow-up is needed to determine if these patients comprise groups at higher risk for recurrence, and therefore benefit from additional chemotherapy or targeted therapies.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 3020.

3-Year Follow-Up of the Partial Breast Irradiation Trial for DCIS Using the MammoSite® Brachytherapy Balloon Catheter

Purpose: To report the 3-year median follow-up of a prospective Phase II partial breast irradiation (PBI) trial utilizing the MammoSite® interstitial balloon as the sole radiation therapy treatment after lumpectomy for ductal carcinoma in situ (DCIS).Materials and Methods: One hundred and thirty-three patients were enrolled from May 2003 through January 2006 to reach the 100-patient partial breast irradiation (PBI) target goal of 34 Gy in 10 fractions, 1 cm peripheral to the MammoSite® balloon surface, b.i.d. with Iridium192 HDR brachytherapy. The eligibility criteria were age 45, unicentric pure DCIS, mammographic size ≤3cm, negative margin of ≥1 mm, gross pathology size ≤5 cm, clinically node negative, post-lumpectomy mammogram absent of residual microcalcificaion. A USC/VNPI score was then calculated based on age, tumor size, margin width, and nuclear grade [ref: Silverstein MJ, Am J Surg 2003;186(4):337-343]. Placement of the balloon was performed at the time of lumpectomy or post-lumpectomy with a required minimum distance of the balloon surface to skin (skin bridge) of 5 mm. Data was collected at enrollment, at implant, 3 and 6-months, then yearly for local control, cosmesis (Harvard Scale), toxicity, adverse events, disease-free survival, cause-specific survival, and contralateral breast failure. Local recurrence is defined as either invasive or non-invasive within the target volume. Ipsilateral elsewhere recurrence (IER) is defined as invasive or non-invasive outside of the target volume.Results: Of the 133 patients enrolled, thirty-three were not treated for the following reasons: less than 5 mm skin bridge (n=13), poor cavity conformance (n=10), positive margin (n=3), microinvasion (n=3), MD decision (n=2), patient request (1), and other (1). The nuclear grade distribution of the tumors were: NG1(17%), NG2(44%), and NG3(39%). The mean age was 60.8 years. The mean tumor size was 10.6 mm; mean closest surgical margin was 6.8 mm (R=0.1-40mm); post-lumpectomy placement in 72%; mean skin bridge distance was 13 mm with 89% ≥7mm. No patients have been lost to follow-up, and at a median 3-year follow-up, the cosmetic results have been rated as excellent/good in 94 and fair in 6 patients. There have been only four recurrences, all non-invasive with the following histological patterns listed in the table below. One was an IER.Recurrence DetailsCase No.Months Since PlacementInvasiveUSC/VNPI ScoreOriginal Tumor GradeComedo Necrosis18No93Yes211No83No317No62Yes432No52Yes No serious adverse events were reported with an infection rate of 9% (7 breast infections; 2 cellulitis).Conclusion: This is the longest reported prospective Phase II study using a PBI technique for pure DCIS patients, and continues to demonstrate the efficacy of the MammoSite® balloon for treating pure DCIS breast tumors, with no new recurrences since our last report at SABCS 2007. All the recurrences were noninvasive, and had at least one or more risk factors of a high USC/VNPI score, high nuclear tumor grade, or had a comedo necrosis pattern. There have been no recurrences in nuclear grade 1 or 2 patients in the absence of a comedo necrosis pattern. The cosmesis also continues to be excellent or good long-term in 96% patients.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 952.

Effect of modest delays in primary surgical treatment on progression of tumor size in breast cancer patients

622

Background: Evaluation of medical co-morbidities, coordination of reconstructive surgery and referral to tertiary care centers can delay surgical treatment in breast cancer. These delays raise concerns for tumor progression in the interim. We evaluated the time from initial imaging at a cancer center to surgical treatment and the change in tumor size. Methods: We identified 823 patients who underwent surgery as their first therapeutic modality for invasive breast cancer diagnosed from December 2003 to September 2005. Baseline tumor size was determined by mammogram (MMG) and ultrasound (US) reports, and tumor size at surgery was determined by pathology reports. Results: The median time from imaging at our facility to surgery was 0.69 months (range 0.03 to 4.34). Multivariate modeling indicated that older patient age, undergoing total mastectomy, and undergoing reconstruction predicted a longer time from initial imaging to surgery. Comparing radiographic to pathologic size, a moderate correlation was demonstrated for both MMG (Spearman r = 0.58; p < 0.0001) and US (Spearman r = 0.66, p < 0.0001). Pathologic size was the same as MMG size in 14%, smaller in 49%, and larger in 37% of patients. Differences in tumor size estimates were not significantly associated with time lapse between MMG and surgery, but in a multivariate model, MMG tumor size, tumor histology, and tumor grade were significant predictors (p < 0.0001 for all) of differences in mammographic and pathologic size. The pathologic tumor size was the same as US in 10%, smaller in 38%, and larger in 52% of patients. Time lapse to surgery was not significantly associated with differences in US and pathologic tumor size. In a multivariate model, US tumor size (p < 0.0001), tumor histology (p = 0.0006) and tumor grade (p = 0.005) were significant predictors of differences in US and pathologic tumor size estimates. Conclusions: There is no evidence that time lapse from initial imaging to surgical intervention leads to significant changes in tumor size thus allowing patients to complete preoperative workup and planning without significant clinical disease progression.

No significant financial relationships to disclose.

The impact of hormone receptor status on pathologic response of HER2-positive breast cancer treated with neoadjuvant chemotherapy with or without trastuzumab

533

Background: The aim of this study was to compare the extent of pathologic response in HER2-positive patients treated with standard neoadjuvant chemotherapy with or without trastuzumab according to hormone receptor (HR) status. Methods: The study included 199 patients with HER2-positive disease treated in clinical trials of neoadjuvant chemotherapy. Eighty-nine patients treated with 3- weekly paclitaxel and concurrent trastuzumab followed by 5-fluorouracil, epirubicin, and cyclophosphamide (FEC) were compared with 110 patients treated with weekly or 3-weekly paclitaxel followed by FEC or FAC. Residual cancer burden (RCB), a measurement of residual disease (RD) from pathologic review of the primary tumor and lymph nodes, was classified as pathologic complete response (pCR), RCB-I (near-pCR, minimal RD), RCB-II (moderate RD), or RCB-III (extensive RD). RCB was compared between treatment groups according to HR status. Results: In HR-negative patients, similar pCR rates were achieved with weekly T/FAC as with 3-weekly T/FEC and trastuzumab (61% and 65%, respectively). However, in HR-positive patients, higher pCR rates were achieved with 3-weekly T/FEC and trastuzumab (47%) than with the weekly T/FAC (25%; p=0.04) or 3-weekly T/FAC (19%; p=0.01) ( Table 1 ). Near pCR (RCB-I) was slightly higher in HR-positive (26%) than in HR-negative patients treated with 3-weekly T/FEC and trastuzumab (11%; p=0.07). Conclusions: Patients with HR- negative/HER2-positive breast cancer had similar pathologic response rates from addition of trastuzumab to 3-weekly T/FEC or from weekly T/FAC chemotherapy. Patients with HR-positive/HER2-positive breast cancer obtained significant benefit from addition of trastuzumab to 3- weekly T/FEC, compared to 3-weekly T/FAC or weekly T/FAC. The combination of weekly T/FEC with trastuzumab as neoadjuvant chemotherapy should be evaluated.

[Table: see text]

[Table: see text]

Two year analysis of treatment efficacy and cosmesis by the American Society of Breast Surgeons (ASBS) MammoSite Breast Brachytherapy Registry Trial in patients treated with accelerated partial breast irradiation (APBI)

529

Background: This report presents 2-year data on treatment efficacy and cosmesis for patients enrolled on the ASBS sponsored MammoSite Registry trial. Methods: 1449 patients were treated with the MammoSite device to deliver APBI. 924 patients have been followed ≥ 12 months, 586 ≥ 18 months, 316 ≥ 24 months and 52 ≥ 36 months. Median follow-up for surviving patients was 14 months. Results: Eleven patients (0.8%) developed an ipsilateral breast tumor recurrence (IBTR) as some component of their initial failure (prior to distant metastases [DM]) for a 2-yr actuarial rate of 1.2%. The 2-yr actuarial rate of isolated IBTR was 0.8% (n=8). Six patients (0.4%) developed an axillary failure (AF) as some component of their initial failure (prior to DM) for a 2-yr actuarial rate of 1.0%. The 2-yr actuarial rate of isolated AF was 0.6% (n=3). The only variable associated with the development of an isolated IBTR (n=8) included an extensive intraductal component (p=0.073) (patient age, margin status, grade, histology, tumor size, nodal status, use of systemic therapy and method of device placement were not associated with IBTR). The only variable associated with the development of all IBTRs (n=11) was tumor location (inner quadrant, p=0.079). The percentage of patients with good/excellent cosmetic results at 6, 12, 18, 24 and 36 months were as follows: 95.1% (n=769 patients), 93.7% (n=621 patients), 91.3% (n=344 patients), 93.5% (n=248 patients), and 90.4% (n=52 patients) (p=NS). For the 174 patients with both a 12-month and 24 month cosmetic assessment, the rate of good/excellent cosmetic results was 94% at 12 and 24 months. At 24 months (n=248 patients), factors associated with good/excellent cosmetic results included increased skin spacing as a categorical variable (94.8% versus 86.1%, p=0.064), no infection (94.8% versus 85.7%, p=0.057), and no systemic chemotherapy treatment (95.3% versus 82.4%, p=0.012). Conclusions: Treatment efficacy and cosmesis 2 years after treatment with APBI using the MammoSite device are excellent and appear similar to those reported with standard whole breast RT.

No significant financial relationships to disclose.

A new measurement of residual cancer burden to predict survival after neoadjuvant chemotherapy

536

Background: The strength of association between tumor response and survival is critical for neoadjuvant chemotherapy trials. Pathologic complete response (pCR) reliably predicts survival benefit, but residual disease contains a range of pathologic responses that likely contain different prognostic groups, including near complete response and resistance. Methods: Pathologic slides and reports were reviewed from 432 patients in two completed neoadjuvant trials: 1) fluorouracil, doxorubicin and cyclophosphamide (FAC) in 189 patients, and 2) paclitaxel followed by FAC (T/FAC) in 243 patients. Paclitaxel was administered as twelve weekly (n=126) or four 3-weekly cycles (n=117). Residual cancer burden (RCB) was calculated as an index that combines pathologic measurements of primary tumor (size and cellularity) and nodal metastases (number and size). We compared four RCB categories, from RCB-0 (pCR) to RCB-3 (chemoresistant), and post-treatment revised AJCC Stage (0-III) for prediction of distant relapse-free survival (DRFS) in multivariate Cox regression analyses (stratified by ER status). Results: The pCR rate was greater after T/FAC than FAC (24% vs. 16%, LR p<0.05), and after weekly (vs. 3-weekly) paclitaxel in T/FAC (30% vs. 16%, LR p<0.01). In patients with residual disease, RCB measurements were significantly lower after T/FAC than FAC (t-test, p<0.0001), but were not different between paclitaxel schedules in T/FAC. RCB was a continuous predictor of DRFS after T/FAC (HR=1.86, 95%CI 1.51–2.30) or FAC (HR=1.67, CI 1.38–2.01) with median follow-up 5 and 8 years, respectively. The resistant category RCB-3 predicted relapse more strongly than AJCC Stage III and identified a larger group of high-risk patients ( Table ). Conclusions: RCB is a new continuous measure of pathologic response that is defined from routine pathologic materials, represents the distribution of residual disease, is a significant predictor of DRFS, and defines chemotherapy resistance more effectively than revised AJCC Stage.

[Table: see text]

[Table: see text]

A single-gene biomarker identifies breast cancers associated with immature cell type and short duration of prior breastfeeding

The pathogenesis of breast cancers that do not express estrogen receptors or Her-2/neu receptors (ER-/HER2- phenotype) is incompletely understood. We had observed markedly elevated gene expression of gamma-aminobutyric acid type A (GABA(A)) receptor subunit pi (GABApi, GABRP) in some breast cancers with ER-/HER2- phenotype. In this study, transcriptional profiles (TxPs) were obtained from 82 primary invasive breast cancers by oligonucleotide microarrays. Real-time reverse transcription-polymerase chain reaction (RT-PCR) was used to measure GABApi gene expression in a separate cohort of 121 invasive breast cancers. GABApi gene expression values from TxP and RT-PCR were standardized and compared with clinicopathologic characteristics in the 203 patients. GABApi gene expression was increased in 16% of breast cancers (13/82 TxP, 20/ 121 RT-PCR), particularly in breast cancers with ER-/HER2- phenotype (60%), and breast cancers with basal-like genomic profile (60%). The profile of genes coexpressed with GABApi in these tumors was consistent with an immature cell type. In multivariate linear regression analysis, the level of GABApi gene expression was associated with ER-/HER2- phenotype (P < 0.0001), younger age at diagnosis (P = 0.0003), and shorter lifetime duration of breastfeeding (< or = 6 months) in all women (P = 0.017) and specifically in parous women (P = 0.013). GABApi gene expression was also associated with combinations of high grade with ER-/HER2- phenotype (P = 0.002), and with Hispanic ethnicity (P = 0.036). GABApi gene expression is increased in breast cancers of immature (undifferentiated) cell type and is significantly associated with shorter lifetime history of breastfeeding and with high-grade breast cancer in Hispanic women.

First analysis of patient demographics, technical reproducibility, cosmesis, and early toxicity

BACKGROUND

Eighty-seven institutions participated in a Registry Trial that was designed to collect data on the clinical use of the MammoSite breast brachytherapy catheter for delivering breast irradiation. Patient demographics, technical reproducibility, cosmesis, and early toxicity were evaluated.

METHODS

From May 4, 2002 through July 30, 2004, 1419 patients with Stage 0, I, or II breast carcinoma who were undergoing breast-conserving therapy were enrolled on the trial. The device was placed in 1403 of these patients. The 1237 patients (87% of enrolled patients) who received accelerated partial breast irradiation (APBI) (34 grays prescribed to 1.0 cm in 10 fractions; 95% of patients who received APBI) constituted the study population; 86% of those patients (1068) had Stages I-II breast carcinoma (median tumor size, 10 mm), and 14% of those patients (169) had Stage 0 breast carcinoma. Ninety-one percent of the patients with invasive carcinoma (977 of 1068 patients) had negative lymph node status, and 99% of all patients had negative margins. The median patient age was 65 years. Systemic chemotherapy alone was administered to 79 patients with invasive carcinoma (7%), hormone therapy was administered to 501 patients (45%), and both were administered to 39 patients (4%). The median follow-up was 5 months.

RESULTS

Five hundred fifty-four catheters (45%) were placed with an open cavity at the time of lumpectomy, and 683 catheters (55%) were placed with a closed cavity after lumpectomy. Skin spacing ranged from 2 mm to 75 mm (median, 10 mm). In 89% of patients, there was a minimum balloon-to-skin distance of 7 mm (2% of patients had distances < 5 mm). In terms of cosmetic assessment, 95% of patients (1030 of 1084 patients) who had a cosmetic assessment had a good/excellent result (last follow-up visit). Cosmetic results at 12 months were good/excellent in 92% of 248 evaluable patients. The median skin spacing (> or = 7 mm vs. < 7 mm) was associated significantly with a good/excellent cosmetic result (96.1% vs. 86.8%; P = 0.0001) overall and at 6 months (P = 0.006). Increasing skin spacing was associated with a good/excellent cosmetic result as a continuous variable (P < 0.0001). In total, 92 of 1140 evaluable patients (8.1%) developed an infection in the breast, which was device-related in 5.3% of patients (60 of 1140 patients). Good/excellent cosmetic results were noted in 86% of these patients (last follow-up visit). Fifteen of 442 evaluable patients (3.4%) developed a radiation recall reaction. Good/excellent cosmetic results were noted in 93% of these patients at their last follow-up visit. One local recurrence (0.1%) was reported (new primary carcinoma).

CONCLUSIONS

Clinical evaluation of the ability of the MammoSite breast brachytherapy catheter to deliver APBI demonstrated acceptable technical reproducibility between multiple institutions and use in appropriate groups of patients. Cosmetic results at 12 months (92% good/excellent) were comparable to those reported with whole-breast RT. Early toxicity rates (infections, radiation recall) appeared to be acceptable.

MRI phenotype is associated with response to doxorubicin and cyclophosphamide neoadjuvant chemotherapy in stage III breast cancer

BACKGROUND

The preferred management for women with stage II or locally advanced breast cancer (LABC) is neoadjuvant chemotherapy. Pathologic response to chemotherapy has been shown to be an excellent predictor of outcome. Surrogates that can predict pathologic response and outcome will fuel future changes in management. Magnetic resonance imaging (MRI) demonstrates that patients with LABC have distinct tumor patterns. We investigated whether or not these patterns predict response to therapy.

METHODS

Thirty-three women who received neoadjuvant doxorubicin and cyclophosphamide chemotherapy for 4 cycles and serial breast MRI scans before and after therapy were evaluated for this study. Response to therapy was measured by change in the longest diameter on the MRI.

RESULTS

Five distinct imaging patterns were identified: circumscribed mass, nodular tissue infiltration diffuse tissue infiltration, patchy enhancement, and septal spread. The likelihood of a partial or complete response as measured by change in longest diameter was 77%, 37.5%, 20%, and 25%, respectively.

CONCLUSIONS

MRI affords three-dimensional characterization of tumors and has revealed distinct patterns of tumor presentation that predict response. A multisite trial is being planned to combine imaging and genetic information in an effort to better understand and predict response and, ultimately, to tailor therapy and direct the use of novel agents.

Functioning pancreatic acinar cell carcinoma: immunohistochemical and ultrastructural analyses

Acinar cell carcinoma (ACC) of the pancreas is a rare malignancy accounting for < 1% of pancreatic neoplasms. We report the clinical and biological characteristics of this carcinoma from two cases that were of interest because of their similar presentation: extensive subcutaneous fat necroses from excessive lipase production by these tumors. Immunohistochemical and ultrastructural analyses of both tumors were consistent with an acinar cell line origin. Recognition of the association between subcutaneous panniculitis and pancreatic neoplasm may prevent long delays in the diagnosis and treatment of this malignancy.