5.20 Impact of Margins on Re-excision Rates for Breast-Conserving Surgery

Posted on January 18, 2018

K. Shuman1, E. Malone2, J. Richman2, C. Parker2  1University Of Alabama at Birmingham,Birmingham, Alabama, USA 2University Of Alabama at Birmingham,Department Of Surgery,Birmingham, Alabama, USA

Introduction:  Approximately 60-75% of breast cancer patients will choose to undergo a lumpectomy. Ideally, an adequate surgical margin is achieved during the initial operation. The definition of an adequate margin has long been debated, which has contributed to the variability in re-excision rates currently ranging from 20-60%. However, in 2014, a new consensus statement was released by the Society of Surgical Oncology (SSO) and the American Society for Radiation Oncology (ASTRO) which defined a negative margin as “no ink on tumor” for stage I and II invasive breast cancers. This retrospective review aims to analyze the impact of the “no ink on tumor” guideline on re-excision rates for patients who have undergone breast conserving surgery at a single institution. We hypothesized that acceptance of this new standardized definition would result in fewer re-excision lumpectomies for patients with stage I and II invasive breast cancers. 

Methods:  We identified all women (≥ 18 years) with a preoperative breast cancer diagnosis who underwent a lumpectomy at a single institution in 2013, the year before the SSO-ASTRO Consensus Statement, and 2015, the year after release.  A re-excision was any procedure documented as an excisional biopsy, biopsy with or without needle localization, or lumpectomy in the medical record of a patient who previously underwent a lumpectomy procedure. Race, age, anesthesia type, and re-excision status were compared. Chi-square tests and t-tests were used to test for bivariate associations between categorical and continuous variables and the year.

Results: Of the 232 malignant lumpectomy cases in 2013, 71 were re-excision surgeries (31%) compared to 64 (24%) of the 268 malignant lumpectomy cases in 2015 (p=0.09). There were no significant differences by age, race, or anesthesia type (all p>0.05).

Conclusion: The release of the SSO-ASTRO consensus statement of “no ink on tumor” has the potential to reduce the amount of additional, unnecessary surgeries for close margins. Reducing re-excision surgeries could improve patient satisfaction and outcomes as these operations pose additional stress on the patient physically, mentally, and economically as well as delay adjuvant therapies. While our data did not reach statistical significance, it suggests a reduction in the number of re-excision surgeries from the year 2013 to 2015, consistent with the expected results of the SSO-ASTRO consensus statement. A larger study will be needed to provide more conclusive evidence.

5.19 Do Disparities Exist Before Breast Cancer Screening: An Analysis of Young Women Without Insurance

Posted on January 18, 2018

E. C. Feliberti1, R. C. Britt1, J. N. Collins1, R. R. Perry1  1Eastern Virginia Medical School,Surgical Oncology,Norfolk, VA, USA

Introduction:
Disparities in breast cancer outcomes can in part be related to access to screening mammography. We hypothesize that these disparities would be minimized in a patient population younger than screening guidelines.

Methods:
Consecutive newly diagnosed sporadic female breast cancer patients under 40 years old treated at an academic medical center were identified and stratified into those without and those with medical insurance. Uninsured were seen in a safety net clinic and offered the same multidisciplinary evaluation and management as the insured counterpart. Patient demographics, tumor histology, treatment rendered and outcomes were compared.

Results:

One hundred twelve patients were identified, 29 without insurance and 83 with insurance. Uninsured women were younger with a median age of 32 y.o. compared to 36 y.o., respectively (p=0.001), with a similar proportion of African American (54.3% vs. 45.3%) and Caucasian (34.3% vs. 48.8%) women. (p=0.292) Median tumor size was 1.85 cm and 2.15 cm, respectively (p=0.312), with similar distribution of luminal A (44.8% vs. 32.7%), luminal B (24.1% vs. 19%), and triple negative (24.1% vs. 41.4%) breast cancers. (p=0.06) The proportion of positive lymph nodes (33% vs 38.5%, p=0.782) and receipt of preoperative chemotherapy (42.9% vs 37.8%, p=0.743) were not different in the 2 cohorts. Uninsured women underwent breast conservation therapy at similar rates as insured (50% vs 45%, p=0.747) and those undergoing mastectomy had similar breast reconstruction rates. (60% vs. 45%, p=0.548) Overall survival of the uninsured and insured cohorts was 86 mos. and 136 mos., (p=0.863) with a respective recurrence-free survival of 75 mos. and not reached. (p=0.885).

Conclusion:
Breast cancer outcomes in those without medical insurance is minimized at younger ages before screening in this single institution study. Outcomes may be more related to tumor biology in this patient population with similar tumor histology, treatment and survival. 
 

5.18 The Impact of Pre-Operative Breast MRI on Surgical Wait Times in a Public Hospital Setting

Posted on January 18, 2018

E. Warnack1, S. Dhage1, K. P. Joseph1  1New York University School Of Medicine,Surgery,New York, NY, USA

Introduction:
Use of MRI for pre-operative evaluation of newly diagnosed breast cancer has become more common, despite questionable impact on survival outcomes. We sought to examine whether or not MRI led to further delay in definitive surgery at this public hospital, and to determine how often and in what manner pre-operative breast MRI changed surgical management. We also sought to examine characteristics of patients who received preoperative MRI.

Methods:
Our breast clinic database was used to identify patients who received surgery between January and December 2015. From this group, patients who received preoperative MRI  were identified. Characteristics of patients, including ethnicity, age, tumor stage, and type of surgery, were collected for both groups. Mean time to surgery, from biopsy definitive operation, was calculated for both groups. Patients who received neoadjuvant chemotherapy were excluded. Of those who received MRI, data on whether MRI changed surgical management was abstracted.

Results:
A total of 101 patients received breast surgery at our institution over a one-year period, and 27 patients received MRI for preoperative planning purposes. There were no significant differences in the MRI and no MRI group in terms of ethnicity (p 0.227.) There were significant differences in the two groups for age, (p .002) stage (p .049,) and type of surgery received (p .005). Patients with stage 2A cancer were 5.1 times more likely (p.026) to receive MRI, and patients with stage 2B cancer were 7 times more likely (p .021) to receive MRI, compared to patients with stage 0 disease. Patients who underwent MRI were less likely to receive lumpectomy or re-excision (OR .212, p .002,) compared to mastectomy. The group of patients who did not undergo MRI experienced slightly longer mean time to surgery (38.75 days compared to 37.4 days in MRI group.)  Of those who received MRI, most (22, 81.4%) had abnormal results, and 13 (48.1%) underwent biopsy as a result of MRI. MRI changed management in nine patients (33.3%,) in most cases by converting a planned lumpectomy to mastectomy. 

Conclusion:

Interestingly, there was no significant difference in time to surgery between the MRI and no MRI group, suggesting that MRI did not cause substantial delay in management. Patients were more likely to receive MRI if they had advanced stages of cancer, and those that received MRI were more likely to receive mastectomy compared to lumpectomy. Considering the high rate of change in surgical management for those who received MRI, and the equivalent time to surgery in this group, it may be inferred that MRI is a helpful imaging study in select patients with breast cancer. Further studies are needed to explore long-term outcomes of those who received MRI. 

 

5.17 The Implications of Insurance Status on Mortality among Early Stage Breast Cancer Patients in Indiana

Posted on January 18, 2018

S. Obeng-Gyasi1, L. Timsina1, K. D. Miller3, G. L. Dunnington1, K. K. Ludwig1, D. A. Haggstrom2  1Indiana University School Of Medicine,Department of Surgery,Indianapolis, IN, USA 2Indiana University School Of Medicine,Department Of Medicine,Indianapolis, IN, USA 3Indiana University School Of Medicine,Division Of Hematology And Oncology,Indianapolis, IN, USA

Introduction: Uninsured and Medicaid insured breast cancer patients have a worse overall survival compared to their privately insured counterparts.  Federally funded programs such as the National Breast and Cervical Cancer Early Detection Program (NBCCEDP) seek to mitigate this disparity by providing uninsured and underinsured women access to preventive care such as screening mammograms. There have been no studies examining whether Indiana’s Breast and Cervical Cancer Program (BCCP) improves treatment outcomes and mortality among breast cancer patients.  To this end, the objective of this study is to identify the differences in overall treatment (surgery, chemotherapy, radiation therapy and hormone therapy) and mortality among uninsured patients diagnosed through Indiana’s BCCP compared to privately insured individuals and government insured individuals (Medicaid or Medicare).

Methods: Study data was obtained using the Indiana state cancer registry and Indiana BCCP records.  Women ages 50-64 with an index diagnosis date of stage 0-III breast cancer  from 01/01/2006-12/31/2013 were included in the study. The data was divided into five insurance groups-uninsured (BCCP only group), privately insured, Medicaid, Medicare and other.  Bivariate intergroup analysis was conducted using chi squared, student T tests, and ANOVA followed by Tukey’s pairwise comparison of the means as appropriate.  Kaplan Meier estimates between the five insurance types were compared using the log rank test.

Results:1477 individuals fulfilled the inclusion criteria.  Approximately 24% of the study population was uninsured, 53% private insurance, 11% Medicaid, 11% Medicare and 1% another payment plan. The groups differed significantly by age, educational attainment, metropolitan status, neighborhood poverty index, clinical stage, surgery type (mastectomy vs lumpectomy), chemotherapy and endocrine therapy. BCCP patients (BCCP 27% vs private 11%) were more likely to present with stage 3 breast cancer (p=0.04).  Intergroup bivariate analysis by stage revealed differences in the utilization of chemotherapy, radiation therapy, surgery type and lymph node surgery based on insurance type. Clinically stage 3 BCCP patients were more likely to undergo a lumpectomy (BCCP 31% vs private 14%, p=0.009) and ALND (BCCP 98% vs private 43%, p=0.001). Kaplan Meier estimates showed uninsured individuals had the highest mortality (p=0.0015).  The divergence in mortality between the uninsured BCCP subjects and the remaining insurance subtypes was evident at 3 years from diagnosis. 

Conclusion: Breast cancer patients diagnosed through Indiana BCCP are more likely to have worse overall mortality. Possible explanations for this disparity include delayed diagnosis or differences in treatment. Future studies should explore how increased access to screening mammography provided by the BCCP program can be better leveraged to realize the goal of reduced disparities in treatment and mortality.

 

5.16 Overview of Florida Genetic Mutation Carriers from the Inherited Cancer Registry (ICARE)

Posted on January 18, 2018

D. A. Henry1, D. Almanza1, C. Lee1, W. Sun1, T. Pal2, C. Laronga1  1Moffitt Cancer Center And Research Institute,Breast Surgical Oncology,Tampa, FL, USA 2Vanderbilt-Ingrahm Cancer Center,Nashville, TN, USA

Introduction: Given the growing list of genetic mutations associated with increased risk for breast cancer, the purpose of this study is to identify Florida enrollees in our Inherited Cancer Registry (ICARE) to better quantify the population for subsequent studies.

Methods: This is a single-institution, IRB-approved international database of subjects recruited from a combination of sources (institutional genetics clinic, external referrals, and social/media outlets) for an inherited cancer syndrome registry. Subjects enrolled from Nov 2000 to Jan 2017 were offered voluntary questionnaires as part of the study. Patients with a Florida zip code and a positive test for one of 11 genetic mutations associated with increased risk for breast cancer (per 2017 NCCN guidelines) were included for analysis. Demographics and questionnaire data were reviewed.

Results: Florida zip codes were identified for 1,247 (55%) of subjects at enrollment. 526/1247 were confirmed carriers of a deleterious mutation, encompassing 8/11 targeted NCCN genes form our study cohort. Median age at enrollment was 48.5 years (range 20-83); 91% were female, 85% self-identified as non-Hispanic Whites. Most patients were BRCA1+ (42%) or BRCA2+ (42%), followed by ATM+ (5%); 374 (71%) of patients had a history of cancer (64% breast; 7% ovarian, 3% both). Questionnaire response rate was 69%. Based on self-reported data, 351/361 (97%) of patients had a high school diploma of which 61% had a college degree or higher; 169/313 (54%) of carriers were the first in their family to have testing.  293/526 (56%) of the cohort, based on zip code, are predicted to fall below the national median household income (MHI) of $54,889; however, where available, questionnaire data placed only 101/314 (32%) below the national MHI.  261 (83%) had private health insurance at the time of genetic testing. 148 (48%) had complete insurance coverage for testing, 98 (32%) received subsidized testing or paid a copay, and 22(7%) paid out of pocket. 161 were treated with surgery for breast cancer; 59 (37%) bilateral mastectomies, 37 (23%) unilateral mastectomies, and 65 (40%) breast-conservation.  72/161 (45%) had genetic testing for surgical decision making, of which 43 (60%) opted for bilateral mastectomy. 89/161 (55%) had genetic testing after surgery, of which 16 (18%) pursued delayed bilateral mastectomy.

Conclusion: Within this educated, insured, and higher income Florida cohort, non-Hispanic White BRCA1/BRCA2 female carriers were the most frequent participants. This is a highly selected group compared to the population of Florida. Genetic test results seemed to influence prophylactic mastectomy choice, as a majority of post-test result cancer patients (59%) elected bilateral mastectomy. This database encompasses a large hereditary cancer syndrome cohort, but is too selected for regional population-based evaluation.

 

5.15 The effect of CTA and venous couplers on surgery duration in microvascular breast reconstruction

Posted on January 18, 2018

L. M. Ngaage1, B. Di Pace2,4, R. Hamed5, G. Oni2, L. Fopp2, B. Koo3, C. M. Malata2,6,7  1University Of Cambridge,School Of Clinical Medicine,Cambridge, , United Kingdom 2Addenbrooke’s Hospital,Plastic And Reconstructive Surgery,Cambridge, , United Kingdom 3Addenbrooke’s Hospital,Radiology,Cambridge, , United Kingdom 4Università Degli Studi Della Campania Luigi Vanvitelli,,Plastic Surgery Unit,Naples, , Italy 5University Of Alexandria,Medical School,Alexandria, , Egypt 6Addenbrooke’s Hospital,Cambridge Breast Unit,Cambridge, , United Kingdom 7Anglia Ruskin University,Faculty Of Medical Sciences,Cambridge & Chelmsford, , United Kingdom

Introduction: Notable amongst recent advances in free flap breast reconstruction, have been the introduction flap imaging technologies (computed tomography angiography (CTA) of the donor vessels) and the use of anastomotic couplers for venous anastomoses. There have been many reports of reduced operative times with the use of CTA or venous couplers (VC). However, none of these reports have compared the effect of these two advances relative to each other. We therefore decided to review the effect of VCs alongside CTA on the operative times of free flap breast reconstruction (FFBR). 

 

Methods:  Following the introduction of venous couplers in June 2010 and the introduction of CTA in November 2011, a retrospective cohort study was conducted of all abdominal FFBRs performed by a single plastic surgeon (CMM) with respect to  CTA and venous coupler use between August 2008 and February 2014. Bipedicled free flaps were excluded (n=26). 40 patients with venous couplers (July 2010 – May 2013) were compared to the 40 patients immediately before the introduction of couplers (August 2008 – June 2010) and 40 patients after the addition of CTA to venous coupler use (November 2011 – February 2014). All CTA patients received venous couplers so it was not possible to compare CTA to venous couplers. Operative time was defined as from knife-to-skin to insertion of the last stitch. Data analysis was completed with SPSS IBM Software. 

 

Results: A total of 120 patients were identified; 40 without intervention (WI), 40 venous couplers only (VC), and 40 with CTA and venous coupler use (CT/VC). The introduction of VC did not significantly reduce the operative time compared to WI (572 vs 586 minutes, p=0.5306). However, patients with both interventions had significantly shorter operative times vs WI (472 vs 586 minutes, p<0.00001). Interestingly, the CT/VC group had a significantly reduced operative time to the VC group (472 vs 572 minutes, p=0.0002), implying that the main factor was the introduction of CTA. Similarly, the use of both modalities reduced the ischaemia time from 100 minutes to 80 minutes (CT/VC vs WI, p<0.00001). CT/VC reduced ischaemia time compared to VC alone (80 vs 89 minutes, p=0.0307). There was also significance between the ischaemia times of WI and VC (100 vs 89 minutes, p=0.0106). 

 

Conclusion: The combined effect of CTA and VC significantly reduced operative and ischaemia times for FFBR; this is predominantly due to use of CTA, although VCs might have a role. The superior effect of CTA on operative time compared to VC is because it facilitates surgery without a surgical learning curve and helps in surgical planning by providing a useful roadmap of the perforator vessels. 

 

5.14 Elevated Body Mass Index is Associated with Early Tissue Expander Removal in Breast Reconstruction

Posted on January 18, 2018

M. Lazar1, C. Silvestri1, I. Le1, S. Weingarten1, I. Chervoneva1, M. Jenkins1, S. Copit1, P. Greaney1, T. Tsangaris1, A. Berger1  1Thomas Jefferson University Hospital,Philadelphia, PA, USA

Introduction: Complications following mastectomy with breast reconstruction can lead to delays in adjuvant therapy for breast cancer.  Common complications include need for extended antibiotics, seroma aspirations, skin necrosis, and early tissue expander removal.  Early removal of a tissue expander is the most devastating.  We sought to identify risk factors for early tissue expander removal. 

Methods: After obtaining IRB approval, we identified patients who underwent mastectomy over a five year period from 1/1/2010 to 12/31/2015.  Ultimately, we found  397 patients who had a mastectomy with reconstruction.  Age, race, body mass index (BMI), as well as history of hypertension, smoking and diabetes were recorded.  The patient charts were reviewed for cancer stage, receipt of neoadjuvant chemotherapy, type of reconstruction, type of mastectomy (unilateral vs. bilateral), and early removal of tissue expander (< 180 days). Univariate and multivariate analyses were performed with p<0.05 being considered significant.

Results: The average age and BMI were 50.5 years old and 28.0. There were 374 patients who underwent mastectomy with tissue expander reconstruction (23 patients had immediate reconstruction with autologous tissue only). 237 patients had expander only reconstruction and 29 of them (12.2%) had early removal. 137 patients had expander + tissue (latissimus flap) and 9 of them (6.6%) had early removal.  This difference between early removal rates was not statistically significant between the two groups (p=0.11). There was a significant difference in BMI between patients with and without early expander removal (30.5 vs. 27.7, p=0.014).  There were 105 patients (28.1%) who underwent neoadjuvant chemotherapy in our cohort; this therapy was not associated with a significant increase in early expander removal (12.4% vs. 9.3%, p=0.45). Additionally, there was no significant difference between early tissue expander removal and mastectomy type (unilateral vs. bilateral, p=0.48), diabetes (p=0.15), hypertension (p=0.56), type of breast cancer (p=0.33), T-stage (p=0.95) and N-stage (p=0.55).  Although it was not statistically significant, the early removal rate was 15.5% in smokers versus 8.6% in non-smokers (p=0.09).

Conclusion: Elevated BMI is a risk factor for early tissue expander removal.  While not statistically significant, smoking may also contribute to early expander removal. These data should be used to counsel patients concerning the risks of surgery.

 

5.13 Contralateral Prophylactic Mastectomy versus Unilateral Mastectomy: Impact on Surveillance

Posted on January 18, 2018

S. R. DeBiase1, W. Sun2, C. Laronga2, D. Boulware3, J. K. Lee3, M. Lee2  1University Of South Florida College Of Medicine,Tampa, FL, USA 2Moffitt Cancer Center And Research Institute,Moffitt Breast Program,Tampa, FL, USA 3Moffitt Cancer Center And Research Institute,Moffitt Biostatistics,Tampa, FL, USA

Introduction:  Contralateral prophylactic mastectomy (CPM) at the time of unilateral breast cancer surgery is increasing. In BRCA+ carriers, CPM reduces contralateral breast cancer risk and is cost-effective, but the cost benefit of CPM in BRCA- patients is controversial. We reviewed breast cancer patients treated with mastectomy and immediate reconstruction; our aim was to evaluate abnormal followup breast imaging and subsequent breast cancers in patients receiving CPM versus unilateral mastectomy (UM) with surveillance.

Methods:  An IRB approved, retrospective, case-controlled, single-institution chart review of breast cancer patients receiving mastectomy and immediate reconstruction from Jan 1990 – May 2013 was performed. Cases were matched 1:1 by reconstruction type and age (+/- 5 years) to limit procedure and age-related confounding variables. Patients with delayed mastectomy, delayed reconstruction, or bilateral cancer diagnosis at surgery were excluded. Staging, pathology, genetic, diagnostic imaging, and outcome data were collected. Therapeutic mastectomy date was used as the reference timepoint. Univariate statistical analyses using SAS (v. 9.4) employed Fisher’s exact test, Wilcoxon Rank Sum, Kruskal Wallis, and Log-Rank tests.

Results: Forty-five UM cases were matched to bilateral mastectomy (BM). Mean age (n = 90) was 52.2 years (range 21.5-74.9) with mean followup time of 7.1 years (range 0.2-19.8).There was no significant difference between UM and BM with regards to BMI, pathologic stage, follow up time, distant recurrence and survival. Genetic status was available for 31.1% of cases; 5 women were BRCA+ and had BM. Six UM and 11 BM patients had abnormal followup breast imaging. Of these, 5 UM and 5 BM patients had abnormal imaging contralateral to the original cancer. Six UM and 10 BM patients had breast/chest wall biopsies after abnormal imaging, with 5 and 4 contralateral biopsies after 4.3 years (range 7.0-0.6) and 4.2 years (range 7.3-1.1) respectively. One UM patient developed contralateral cancer; 5 BM patients had local recurrence. Two UM and 5 BM patients had distant recurrences. Twenty-one UM and 16 BM patients complained of contour/asymmetry after mastectomy, 17 UM and 26 BM did not;  7 UM and 3 BM patients did not report cosmesis (P=0.18, Fisher’s Exact test).

Conclusion: Our results demonstrate the low frequency of contralateral abnormal imaging in UM patients and extremely low incidence of contralateral breast cancer with close followup. More patients with CPM had new abnormal imaging and biopsies, although this was not statistically significant. Thus, CPM did not reduce contralateral imaging/biopsies and additional biopsies may actually be related to CPM with reconstruction.
 

5.12 Breast Cancer in Women over 80: An Analysis of Treatment Patterns and Disease Outcomes

Posted on January 18, 2018

E. N. Ferrigni1,2, C. Bergom1,2, Z. Yin1, A. Kong1,2  1Medical College Of Wisconsin,Milwaukee, WI, USA 2Froedtert Hospital,Surgical Oncology,Milwaukee, WI, USA

Introduction:  With a growing female octogenarian population due to improvements in medical care, there is an increasing number of elderly women diagnosed with breast cancer. No clear standard treatment guidelines exist for older women with invasive breast cancer, and few breast cancer clinical trials include elderly women or examine the impact treatment has on survival. This study aims to examine the practice patterns and treatment outcomes of elderly women ≥80 years old with invasive breast cancer.

Methods: From 2005-2014, female patients ≥80 years old with Stage I-III invasive breast cancer were identified in our hospital tumor registry. Treatments assessed included surgery, radiation, and systemic therapies. Co-morbidities were assessed via the Charlson Comorbidity Index. Descriptive statistics were performed for continuous variables and categorical variables. Kruskal-Wallis and Chi-square tests were applied to evaluate the associated continuous variables and categorical variables, respectively. Kaplan-Meier plot with log-rank test was performed to evaluate survival.

Results: The final cohort consisted of 124 patients. Median age of diagnosis was 84 (range 80-99). 48% of cancers were detected by mammography. 90% of tumors were ER+, 78% were PR+, and 11% were HER2+. 90% of patients underwent surgery (112/124), and those patients were younger (p<0.001). For those who had surgery, 73% underwent a lumpectomy while 27% received a mastectomy. There was no difference in comorbidities between the surgical and non-surgical group (p=0.817). Only 13% of the patients received chemotherapy, whereas 74% and 45% received hormonal therapy and radiation, respectively. Of those who received surgery, 20 patients (12%) experienced complications within 3 months of their procedure. There were 14 cases with reported radiation and/or chemotherapy complications within 3 months. On multivariate analysis, patients greater than the age of 87 (p=0.005) were less likely to have surgery. Overall survival probability (p=0.01) and event-free survival probability (p=0.05) was significant and borderline significant, respectively, in those who received surgery compared to those who did not.

Conclusion: In this cohort, patients who were younger, with image-detected tumors were more likely to undergo surgery. Given the low frequency of complications with surgery, radiation, or chemotherapy, including these therapies as standard of care for octogenarians who have few comorbidities and reasonable longevity should be considered. Further analysis of treatment outcomes in this patient population is warranted to determine the impact of treatment on overall survival.

 

5.11 The role of CTA in assessing DIEP flap perforator patency in patients with pre-existing abdominal scars.

Posted on January 18, 2018

L. M. Ngaage1, R. Hamed3, B. Di Pace2,5, G. Oni2, B. Koo4, C. M. Malata2,6,7  1University Of Cambridge,Clinical School Of Medicine,Cambridge, ENGLAND, United Kingdom 2Addenbrooke’s Hospital,Plastic & Reconstructive Surgery,Cambridge, ENGLAND, United Kingdom 3The University Of Alexandria Medical School,Alexandria, ALEXANDRIA, Egypt 4Addenbrooke’s Hospital,Radiology,Cambridge, ENGLAND, United Kingdom 5Università Degli Studi Della Campania Luigi Vanvitelli,Plastic Surgery Unit,Naples, NAPLES, Italy 6Addenbrooke’s Hospital,Cambridge Breast Unit,Cambridge, ENGLAND, United Kingdom 7Anglia Ruskin University,Postgraduate Medical Institute, Faculty Of Medical Sciences,Cambridge & Chelmsford, ENGLAND, United Kingdom

Introduction: Abdominal scars can affect the patency of deep inferior epigastric artery (DIEA) perforators and are a concern when planning abdominal free flap breast reconstruction (FFBR). Computed tomographic angiography (CTA) is routinely used for preoperative DIEA flap imaging. Few studies address the effects of scars on DIEA perforator patency. We, therefore, investigated CTA utility in predicting the most clinically useful DIEA perforators in scarred abdomens.

 

Methods: A single surgeon’s first 100 CTA FFBR patients were studied. All were imaged by one radiologist (BK). CTA reports, abdominal scars and flap intraoperative details were analysed. The operative findings were then correlated with the CTA "predictions".

 

Results: 100 patients with preoperative CTAs underwent 141 FFBRs 39% (55) from scarred and 61% (86) from virgin abdomens. All flap transfers were successful. Overall, concordance between the best perforator chosen by CTA pre-operatively and that selected by the surgeon intraoperatively was 95.6% (scarred 92.7%; non-scarred 97.5%,?p=0.18). There was no difference in the proportion of single-perforator flaps in the two groups (scarred 39%; non-scarred 37%). “Scarred” flaps were heavier (775 vs 675g, p=0.04) and their reconstructions took an hour longer (563 vs 502 minutes, p=0.03). Two patients had abnormal incidental CTA findings that precluded flap harvest (DIEA occlusion from previous surgery and a large mesenteric aneurysm).

 

Conclusion: CTA accurately predicted perforator choice in flaps from scarred and virgin abdomens. Scarring increases the duration of FFBR. Discovery of incidental CTA abnormalities can prevent doomed-to-fail or unsafe free flap surgery. Our study confirms the utility of CTA in facilitating FFBR from scarred abdomens.? 

 

5.10 The Effect of Obesity on Operating Room Utilization in Breast Surgery

Posted on January 18, 2018

N. Tata1, J. Dunderdale2, I. Helenowski3, B. Jovanovic3, R. Marcus4, S. Kulkarni2  1Feinberg School Of Medicine – Northwestern University,Chicago, IL, USA 2Feinberg School Of Medicine – Northwestern University,Department Of Surgery,Chicago, IL, USA 3Feinberg School Of Medicine – Northwestern University,Department Of Preventative Medicine,Chicago, IL, USA 4Feinberg School Of Medicine – Northwestern University,Department Of Anesthesia,Chicago, IL, USA

Introduction: The current obesity epidemic is associated with increased health care costs and comorbidities such as diabetes and heart disease.  However, the effect of obesity on OR utilization has not been completely evaluated.  The goal of our study was to examine how operative time (OPT) and total time in the operating room (TTOR) for common breast procedures are affected by patient BMI.  We hypothesized that operating room utilization would vary significantly with BMI, with the difference being greater for more extensive breast procedures compared to minimally invasive procedures. 

Methods:  For this study, the American College of Surgeons National Surgical Quality Improvement Project (NSQIP) databases for 2010-2012 were searched.  Patients undergoing the selected breast operations were filtered out by CPT code.  They were divided into groups based on their BMI and their weight.  Using the two-sample t-test, OPT and TTOR for the procedures were compared among the lowest and highest BMI categories, as well as the lowest and highest weight categories. To further characterize the effect of BMI on operative time, a linear regression t-test was used to demonstrate increase in OPT as a function of unit increase in BMI.

Results:  When the lowest and highest BMI groups were compared for all procedures, significant differences in OPT and TTOR were seen (p<0.0001).  Ultimately, our analysis included 47,557 patients for OPT data and 32, 455 for TTOR data.  Overall, there was a fourteen minute difference in OPT and an eighteen minute difference in TTOR.  Similarly, when the lowest and highest weight categories were compared for all procedures, a significant difference in OPT of thirteen minutes was seen, while the difference in TTOR was seventeen minutes.  In both BMI and weight analyses, though significant differences were noted for lumpectomy alone and lumpectomy plus SLNB, the effect of patient BMI on ALND and mastectomy is even more pronounced.

For every ten unit increase in BMI, there was a 9.6 minute increase in operative time for lumpectomy ALND and 14.3 minutes increase for mastectomy ALND procedures.  For simpler procedures like lumpectomy and lumpectomy SNLB, the increase in operative time with ten unit BMI increase was 2.4 and 5.2 minutes respectively.

Conclusion: Patient BMI and weight significantly affect OPT and TTOR for common breast procedures.  However, the difference is greater for ALND and mastectomy.  Therefore, when scheduling more extensive breast procedures, patient BMI should be taken into account to improve operating room scheduling and adjust physician compensation.  

5.09 The Effect of Surgeon Performed Intra-Operative Specimen Ink on Lumpectomy Re-excision Rates

Posted on January 18, 2018

A. Botty Van Den Bruele1, B. Jasra1, C. Smotherman2, M. Crandall1, L. Samiian1  1University Of Florida College Of Medicine Jacksonville,Department Of Surgery,Jacksonville, FL, USA 2University Of Florida College Of Medicine Jacksonville,Center For Health Equity And Quality Research,Jacksonville, FL, USA

Introduction: A key factor in breast conservation therapy is obtaining negative surgical margins. Historical positive margin rates have been 20-40%, requiring re-excision to obtain clear margins.   Recent ASTRO-SSO margin guidelines define a negative margin as no tumor at the inked margin for invasive cancer, and less than 2mm from the ink in DCIS.  Discordance between the surgeon and the pathologist interpretation of specimen orientation has been reported to be as high as 31% and could influence the accuracy of re-excisions. This study examined whether the addition of surgeon performed intraoperative inking of the specimen would reduce re-excision rates after initial lumpectomy for breast cancer.

Methods: A retrospective review of a single institution prospective surgical database was performed from August 2009- May 2017 and included patients who had initial lumpectomy with pre-op diagnosis of invasive breast carcinoma or DCIS. Intraoperative specimen inking of all initial lumpectomy specimens was performed by the surgeons after Nov 2015. Re-excision rates after initial lumpectomy was compared across three time periods: before margin guideline publication (Jan 2014) vs. after guideline adoption (Jan 2014 – Oct 2015), vs. after the addition of surgeon performed intraoperative specimen ink (Nov 2015- May 2017).

Results: A total of 400 initial lumpectomies for DCIS and invasive carcinoma were evaluated. Overall re-excision rate was 21% (n=84). There was no difference in overall re-excision rate across the 3 time periods.  Patient with DCIS were 2.8 times more likely to undergo re-excision for margins as compared to patients with invasive carcinoma (p<.0001) and this difference persisted across all time periods.  There was a consistent reduction in re-excision for invasive cancer with adoption of new guidelines and addtion of IOP specimen ink.

Conclusion: Re-excision rates after initial lumpectomy remain significantly higher for DCIS than for invasive disease. Although margin guidelines improved re-excision rates, the addition of surgeon performed intraoperative inking of the lumpectomy specimen provided added reduction in re-excision of invasive carcinoma, but not in DCIS.  Better understanding of biology of DCIS may improve local therapy for this entity.

5.08 Retrospective Analysis of Acellular Dermal Matrix Efficacy following Immediate Breast Reconstruction

Posted on January 18, 2018

N. Sobti1, E. Ji1, R. L. Brown1, C. L. Cetrulo1, A. S. Colwell1, J. M. Winograd1, W. G. Austen1, E. C. Liao1  1Massachusetts General Hospital,Division Of Plastic And Reconstructive Surgery,Boston, MA, USA

Introduction:  The number of mastectomy procedures has risen 5.1% within the past decade, where nearly one-third of patients undergo immediate breast reconstruction. Prosthesis-based reconstruction using a staged approach, where tissue expander (TE) is placed at the time of mastectomy followed by tissue expansion and subsequent operation for exchange for implant, is the most common method of breast reconstruction. Over the last decade, TE placement has been increasingly performed utilizing acellular dermal matrix (ADM), such that 60% of alloplastic breast reconstructions are ADM-based. ADM is a de-cellularized cadaveric soft tissue graft that provides structural support and controls prosthesis positioning. Although numerous studies have reported safety outcomes of breast reconstruction using ADM, this study uniquely evaluates the efficacy of ADM as soft tissue reinforcement by examining the objective parameters of initial fill volume at the time of mastectomy, number of expander fill, and time duration between exchange of TE for final implant.

Methods:  Retrospective chart review of immediate breast reconstruction cases with TE was conducted at a tertiary academic medical center over 12 years. Procedures were performed by placing TE in the sub-pectoral position, with either ADM or serratus and myo-fascial flaps (non-ADM) to support the implant in the lower pole of the breast. Univariate and binomial regression analyses were performed to compare endpoints between ADM and non-ADM groups.

Results: Patients who underwent ADM-based breast reconstruction achieved a significantly higher initial TE fill volume when compared to those in the non-ADM cohort (180.8 ± 150.0 v. 45.8 ± 74.4 respectively, p = 0.00). Normalizing for final implant size, the ADM group exhibited significantly higher perioperative fill volume than the non-ADM group (0.33 ± 0.24 v. 0.11 ± 0.16, p = 0.00). Additionally, the ADM cohort experienced fewer TE expansion visits (5.0 ± 2.0 fills v. 6.7 ± 2.6 fills, p = 0.00) and shorter time interval between TE placement and implant exchange (5.4 ± 3.6 months v. 7.0 ± 4.9 months, p = 0.00). Complication rates were comparable between groups. In addition, we observed a collinear trend between ADM use and direct-to-implant single stage procedures during the study period.

Conclusion: This study demonstrated efficacy of ADM in immediate TE-mediated breast reconstruction. These results suggest that ADM-based procedures are associated with greater perioperative fill volume and shorter time duration between TE placement and implant exchange. Furthermore, the increased use of ADM correlated with the rise of single-stage breast reconstruction at our institution, thereby removing need for expansion and second stage implant exchange procedure. This work serves as a framework for future studies evaluating ADM efficacy in breast reconstruction and can guide development of emerging biomaterials and techniques.

 

5.07 Recurrence in Patients Who Achieved Pathological Complete Response by Neoadjuvant Chemotherapy

Posted on January 18, 2018

M. Asaoka1, K. Narui3, A. Yamada3, N. Suganuma4, T. Chishima5, K. Takabe2, T. Ishikawa1  1Tokyo Medical University Hospital,Department Of Breast Cancer,Tokyo, TOKYO, Japan 2Roswell Park Cancer Institute,Buffalo, NY, USA 3Yokohama City University Hospital,Department Of Breast Cancer,Yokohama, KANAGAWA, Japan 4Kanagawa Cancer Center,Department Of Mammary Gland Endocrine Surgery,Yokohama, KANAGAWA, Japan 5Yokohama Rosai Hospital,Department Of Oncology,Yokohama, KANAGAWA, Japan

Introduction:

Recent studies have indicated that patients who achieved a pathological complete response (pCR) by neoadjuvant chemotherapy (NAC) have better long-term outcomes than those who did not. Recently, the pCR rate is approaching to 50% particularly in patients with hormone-receptor negative disease. If the disease is not recurred locally in cases with pCR and pCR could be accurately diagnosed preoperatively, it may be possible to treat some population of patients without surgery after chemotherapy. We analyzed the outcomes of patients who had achieved pCR by NAC with a special attention to local recurrence and risk factors of recurrence.

Methods:

We investigated disease free survival in 395 patients who were identified as having a pCR from 1599 patients with primary operable breast cancer treated by NAC in 4 institutions (pCR rate of 24.7%; 395/1599). As for subtypes in 395 cases, pCR cases were 50 in Luminal type (pCR rate of 7.2%), 98 in Luminal-HER2 type (32.1%), 116 patients in HER2 type (52.5%), and 131 in triple negative (TN) type (34.2%). 

Results:

The median follow-up was 41 months. Recurrent diseases including local recurrence or distant metastasis was found in 5.80% (23/395). According to subtypes, these were 2.00% (1/50) in Luminal type, 4.08% (4/98) in Luminal-HER2 type, 10.3% (12/116) in HER2 type, and 4.58% (6/131) in TN type. Local recurrence was found in 1.2% of all cases (5/395). It was prominent that brain metastasis was frequently observed in HER2 type (12/116). Clinical stage before NAC and nodal status after NAC were found as a risk factor of recurrence in the univariate analysis, and only clinical stage remained statistically significant in the multivariate analysis.

Conclusion:

Except HER2 type, recurrence was not frequent in cases obtained pCR, particularly in cases with an early clinical stage. Local recurrence was rarely observed in any subtype.  Based on this result, we think that it is possible to omit surgery in patients highly expected pCR. We have already conducted a multicenter feasibility study to treat without surgery. For cases diagnosed as clinical complete response after NAC by contrast-enhanced magnetic resonance imaging, ultrasound-guided core needle biopsy (CNB) is performed before starting the surgery. The concordance of pathological results between CNB and surgical specimen is examined. The enrollment was completed lately.

 

5.06 Effect of Preoperative MRI on Rate of Local and Distant Recurrence of Breast Cancer

Posted on January 18, 2018

A. Roy1, Z. Zeng3, X. Li6, S. Espino4, Y. Luo3, H. Jiang5, S. Khan4  1Northwestern University,Feinberg School Of Medicine,Chicago, IL, USA 3Northwestern University,Department Of Preventative Medicine, Feinberg School Of Medicine,Chicago, IL, USA 4Northwestern University,Department Of Surgery, Feinberg School Of Medicine,Chicago, IL, USA 5Northwestern University,Department Of Statistics,Evanston, IL, USA 6Harvard School Of Public Health,Department Of Social & Behavioral Sciences,Boston, MA, USA

Introduction: The most sensitive imaging modality currently used to detect multifocal and metacentric breast cancer is magnetic resonance imaging (MRI), yet its role in preoperative evaluation of disease extent remains controversial. It was initially hoped that preoperative MRI use would allow more complete resection of disease and would improve outcomes, but a prospective study has shown that it does not reduce re-excisions and retrospective analyses do not support an improvement in cancer outcomes. However, the number of local recurrences and patients in these studies is small. In an attempt to clarify the potential benefit of pre-operative breast MRI for long-term breast cancer outcomes, we report a retrospective review of data on 3902 women diagnosed with primary breast cancer at the Lynn Sage Breast Center of Northwestern Medicine.  

Methods: The Enterprise Data Warehouse of Northwestern Medicine was searched for women diagnosed with ductal carcinoma in situ (DCIS) or invasive breast cancer who underwent breast conservation therapy (BCT) between 2000-2016. The use of preoperative MRI was extracted along with clinical and therapeutic details. The frequencies of local recurrence (LR) and distant recurrence (DR) were evaluated with Cox proportional hazards model, adjusting for age, race, tumor size, tumor grade, lymph node status, ER status, PR status, HER2 status, P53 status, Ki67 status, systemic therapy status, and radiation therapy status.  

Results: Among 3902 women with primary breast cancer, 1,303 had preoperative MRI and 2,599 did not. Compared to the women who did not have MRI, women with MRI were younger (55 vs 59 years, p<0.0001), had larger tumor size (1.64 cm vs 1.55 cm, p=0.03), and underwent systemic therapy more frequently (p<0.0001). Median follow-up time for the MRI group was 75 months, and for the non-MRI group was 125 months (p<.0001). Ipsilateral LR was experienced by 224 women (5.74%), and DR occurred in 227 women (5.82%). In univariable Cox regression models, the hazard ratio (HR) with use of MRI was 0.94 (95% CI 0.71 to 1.24; p=0.65) for LR; and 0.84 (95% CI 0.62 to 1.15; p=0.29) for DR. In multivariable Cox regression models, the HR with use of MRI was 0.88 (95% CI 0.65 to 1.17; p=0.37) for LR; and 0.77 (95% CI 0.56 to 1.06; p=0.11) for DR.   

Conclusion: Women who received preoperative MRI differed significantly from those who did not, but cancer outcomes for either local or distant recurrence following BCT were not significantly different by MRI use. However, this and other retrospective analyses are likely subject to bias given the factors that appear to drive the use of MRI.   

 

5.05 Expression of miR-9 Predicts Breast Cancer Survival

Posted on January 18, 2018

J. C. Sporn1, E. Katsuta1, L. Yan2, K. Takabe1  1Roswell Park Cancer Institute,Surgical Oncology,Buffalo, NY, USA 2Roswell Park Cancer Institute,Biostatistics & Bioinformatics,Buffalo, NY, USA

Introduction:

miRNAs are a diverse family of RNA molecules. They are typically about 18 to 24 nucleotides in length and regulate translation and stability of partially complementary target mRNAs. It has been shown that miRNA expression is severely dysregulated in cancer cells and that altered expression of certain miRNAs can promote tumorigenesis. Some miRNAs were found to be overexpressed in cancer cells and shown to promote proliferation in vitro and tumorigenesis in vivo, other miRNAs may be downregulated in cancer cells subsequently failing to inhibit the expression of oncogenic protein and thus promoting tumorigenesis. miR-9 was shown to increase breast cancer cell motility and invasiveness in vitro and to promote metastasis formation in mice by inhibition of E-cadherin expression. Interestingly, miR-9 was also linked to downregulation of two sirtuins, SIRT1 and SIRT3. Sirtuins (of which there are seven in humans) are NAD+-dependent enzymes involved in a variety of cellular pathways including stress-response and chromatin-silencing. SIRT1 is downregulated in BRCA1-associated breast cancer compared with normal controls and inhibition of BRCA1-mutant tumor growth by the anticancer agent resveratrol upregulated SIRT1 activity.

Methods:

We utilized the TCGA dataset to analyze the expression of miR-9-5p in human breast cancer patients. To link the expression levels with the provided TCGA survival data, we used the OncoLnc platform. 988 patients were divided at the 50th percentile into two equal groups of 494 according to their expression levels (low expression versus high expression). Cox regression was performed and a Kaplan plot was calculated. We further compared the expression levels of SIRT1 to SIRT7 between the groups with low and high expression of miR-9-5p.

Results:

The miR-9-5p expression ranged from 22.98 to 562.5 in the ‘low’ and from 567.51 to 164644.95 in the ‘high’ expression group. Survival analysis showed that high expression of miR-9-5p was associated with worse prognosis (p-value= 0.00801). This is in line with previous in vitro findings linking increased miR-9 expression with invasiveness and metastatic potential in breast cancer. While there was a trend towards lower expression of SIRT1 in the group with high expression of miR-9-5p, it did not reach statistical significance. However, there was a statistically significant decrease in SIRT3 levels in the group with high expression of miR-9-5p (p<0.001).

Conclusion:

Our analysis supports an important link between miR-9 and sirtuins and identifies miR9-5p as a novel biomarker to predict survival in breast cancer patient.

5.04 Analysis of neoadjuvant chemotherapy for invasive lobular carcinoma

Posted on January 18, 2018

L. A. Riba1, T. L. Russell1, T. A. James1  1Beth Israel Deaconess Medical Center,Surgery,Boston, MA, USA

Introduction:  The achievement of pathological complete response (pCR) following neoadjuvant chemotherapy (NCT) for breast cancer is associated with improved survival and enables the use of more conservative surgical management. Invasive lobular carcinoma (ILC) has been extensively shown to have a lower response rate to NCT than invasive ductal carcinoma (IDC). The purpose of this study is to summarize the national trends in the use of NCT for ILC, to characterize the overall pCR in ILC patients compared to IDC patients, and to determine if there is a subset of patients with ILC who show improved pCR rates. 

Methods:  This is a retrospective study using data from the National Cancer Database. The population consisted of female patients with clinical T1-T4 primary ILC or IDC diagnosed and treated between 2010 and 2014. For patients receiving NCT, the event of pCR was assessed by using the variable ‘Response to Neoadjuvant Therapy’, and each patient was categorized as either presenting with pCR or not. Clinically relevant variables, including age, TNM stage, tumor grade and tumor receptor status, were used in multivariate logistic regression models for each histologic subtype, modeling for the event of pCR. From the multivariate regression model we obtained odds ratios and 95% confidence intervals (CI) for each variable as a predictor for pCR in either ILC or IDC cases.

Results: Our study population consisted of 387,200 women, of which 89.47% had IDC and 10.53% had ILC. Patterns of systemic therapy varied significantly between the two histologic subtypes with NCT being used in 15.39% of the IDC cases and only in 8.57% of the ILC cases. A significantly higher rate of pCR was found in the cases of IDC compared to those of ILC (22.8% compared to 8.08%; p < 0.0001). Multivariate logistic regression analysis found a significant relationship between increased rates of pCR in lobular carcinoma with HER2-positive subtypes (OR, 11.309; 95% CI, 6.812-18.777) and TNBC cases (OR, 4.484; 95% CI, 2.924-6.876). Patients with older age (OR, 0.273; 95% CI, 0.075-0.999) were seen to have significantly lower response rates in the ILC group. 

Conclusion: From a treatment response standpoint, the use of NCT in women with ILC has not been proven widely beneficial. While complete response to NCT in patients with ILC is unusual, patients with HER2 positive and triple negative disease demonstrate pCR rates comparable to those found in cases of IDC. Our findings demonstrate selective benefit from NCT in the treatment of ILC, specifically restricted to those patients presenting with HER2-positive tumors or TNBC.

 

5.03 Bone Morphogenetic Protein Expression Significantly Affects Breast Cancer Prognosis

Posted on January 18, 2018

A. A. MAAWY1, E. Katsuta1, L. Yan2, K. Takabe1  1Roswell Park Cancer Institute,Surgical Oncology,Buffalo, NY, USA 2Roswell Park Cancer Institute,Biostatistics And Bioinformatics,Buffalo, NY, USA

Introduction:
Bone morphogenetic proteins are members of the TGFβ family of signaling pathways and are known to be essential in fetal development, tissue differentiation and a multitude of cellular functions. Differential expression noted in some forms of breast cancer and is known to regulate the epithelial to mesenchymal transition, tissue infiltration and metastasis. Under-expression of BMP7 is known to be associated with a more metastatic phenotype and development of bone metastases. Studies in these patients have revealed aberrations of both BMP expression and signaling, which correlate clinically with disease progression. Identifying these tumor subsets can act as prognostic markers and may prove to be a therapeutic target with implications on disease progression and overall prognosis. This study investigates the association of BMP gene expression with breast cancer survival using a ‘big data’ approach employing RNA sequencing from the Cancer Genome Atlas (TCGA).

Methods:
A total of 1096 patient with breast cancer had treatment naïve samples of the tumors undergo genetic sequencing and the results of their sequencing stored in The Cancer Genome Atlas (TCGA) dataset. Overall survival (OS) was compared between high and low expression of indicated BMP related genes; BMP1, BMP2, BMP3, BMP4, BMP5, BMP6, BMP7, BMP8a, BMP8b, BMP10, BMP15 and BMP receptors 1A and 1B, based upon RNA-sequencing data of TCGA.

Results:

The TCGA cohort was representative of national breast cancer patients with respect to stage, pathology and survival. Using Kaplan Meier analysis, BMP1 (p<0.001), BMP3 (p=0.002), BMP5 (p=0.02), BMP6 (p<0.001), BMP7 (p<0.001), BMP8a (p=0.054), BMP8b (p=0.001), BMPR1A (<0.001)and BMPR1B (p=0.005) all significantly impacted OS. BMP2, BMP4, BMP10 and BMP15 did not significantly impact OS. High expression of BMP1, BMP2, BMP3, BMP4, BMP5, BMP7 and BMPR1A was protective and was associated with improved OS. Conversely, high expression of BMP6, BMP8a, BMP8b and BMPR1B was associated with poorer outcomes and worse OS.

Conclusion:

BMP expression profiles may be of value in prognostication. Evidence suggests that BMPs play a role in breast tumorigenesis, function and progression by modulating the cell cycle, interactions with and remodeling of the extracellular matrix, interactions with the bone metastatic microenvironment and the epithelial to mesenchymal transition. Intervention in this pathway may serve to improve outcomes, manage metastatic disease and assist in clinical decision making on optimal therapy based on risk of recurrence or metastasis.

5.02 THE EFFECT OF CONTRALATERAL PROPHYLACTIC MASTECTOMY (CPM) ON COMPLICATIONS AND LENGTH OF STAY (LOS)

Posted on January 18, 2018

J. Huang1, A. Chagpar1  1Yale University School Of Medicine,New Haven, CT, USA

Introduction:  The rate of CPM has been increasing amongst patients with unilateral breast cancer. There is a paucity of data, however, regarding the impact of CPM on complications requiring reoperation within 90 days, and length of hospital stay (LOS).

Methods:  Medical records of female patients presenting with unilateral breast cancer who underwent a mastectomy at a large academic institution between July 2013-July 2016 were analyzed to determine the effect of CPM on complications requiring reoperation within 90 days of surgery, and LOS.  Non-parametric statistical analyses were performed using SPSS version 24.

Results:  471 patients were included in this cohort; 276 (58.6%) opted for CPM. Median patient age was 52 (range; 22-90). Patients opting for CPM tended to be younger (median age 48 vs. 61, p<0.001), have a BRCA mutation (13.0% vs. 1.5%, p<0.001), have smaller invasive tumors (0.8 cm vs. 1.6 cm, p<0.001), have private insurance (74.6 vs. 63.6%, p<0.001), have had neoadjuvant chemotherapy (CTx, 33.0% vs. 19.1%, p=0.001), and be more likely to opt for reconstruction (94.2% vs. 57.4%, p<0.001) than those who had unilateral mastectomy (UM). CPM patients were more likely to be white or black; less likely to be Asian (p=0.025). Body mass index (p=0.653), smoking status (p=0.746), receipt of adjuvant CTx (p=0.082), receipt of radiation therapy (p=0.482), and history of diabetes (p=0.294) were not significantly different between the two groups. 52 patients (11.0%) had a complication within 90 days requiring reoperation; however, the rate of complications among CPM patients was no different than that among UM patients (11.2% vs. 10.8%, p=1.000). Median LOS for the overall cohort was 2 days. CPM patients tended to have longer LOS than UM patients (median 3 vs. 2 days, p<0.001); however, their rate complications within the initial hospital stay was the same as for UM patients (3.6% vs. 4.1%, p=0.811). Controlling for factors that were associated with CPM on bivariate analysis, CPM remained a significant independent predictor of LOS ≥ 2 days (OR=2.369; 95% CI: 1.197-4.688, p=0.013). Older age (OR=1.041; 95% CI: 1.009-1.075, p=0.012) and reconstruction (p<0.001) were also predictors of longer LOS.  Compared to those who did not have reconstruction, those who opted for implant based reconstruction (OR=6.791; 95% CI: 3.247-14.201, p<0.001) and those who opted for flap based reconstruction (OR=196.522; 95% CI: 42.315-912.708, p<0.001) were more likely to have a LOS ≥ 2 days, controlling for other factors.

Conclusion:  Over 50% of patients with unilateral breast cancer who underwent mastectomy opted for CPM. CPM is associated with an increased LOS, but this is not due to a higher complication rate, and there is no association between CPM and 90 day reoperation rate. These data suggest that the risk of greater complications should not be a significant detractor in decision-making regarding CPM.

5.01 Identifying low risk populations for the omission of sentinel lymph node biopsy in breast cancer

Posted on January 18, 2018

L. A. Riba1, T. A. James1  1Beth Israel Deaconess Medical Center,Surgery,Boston, MA, USA

Introduction:  Sentinel lymph node biopsy (SLNB) is recognized as a safe and efficient method for axillary node evaluation in clinically node negative patients with breast cancer. Although developed as a less invasive alternative to axillary lymph node dissection (ALND), SLNB is not free of risks, added costs and patient discomfort. Furthermore, data have already begun to identify subsets of low-risk patients where SLNB may not be required (i.e. women over 70 with hormone receptor positive breast cancer). The purpose of this study was to determine if there were additional low risk subsets were this procedure may be safely omitted because of a low rate of axillary node involvement. 

Methods:  This is a retrospective study using data from the National Cancer Database. The population consisted of female patients with pathological T1-T2 primary invasive ductal carcinoma (IDC) who underwent a SLNB between 2012 and 2014. Descriptive statistics were used to analyze the main characteristics of the study population and determine the rate of node positivity related to each clinically relevant characteristic. These clinically relevant characteristics, including age group, ethnicity, tumor grade, tumor size (using pathologic TNM classification) and tumor receptor status, were then used in a multivariate logistic regression, modeling for negative nodes following SLNB. From the multivariate regression model we obtained odds ratios and 95% confidence intervals for each variable as an indicator of negative SLNB.

Results: Our study population consisted of 114,916 women who underwent SLNB, of which 79.37% had negative nodes and 20.63% had positive nodes. Multivariate logistic regression analysis found a significant relationship between a negative SLNB result and older age, Asian ethnicity, tumor size smaller than 20mm and triple negative tumor receptor status. Small tumor size was found to be the strongest indicator of negative SLNB, particularly microinvasive tumors (OR, 8.29; 92% negative SLNB), pT1A tumors (OR, 11.11; 94% negative SLNB), pT1B tumors (OR, 6.28; 91% negative SLNB), and pT1C tumors (OR 2.36; 79% negative SLNB). Older age was also associated with increased negative SLNB rates, with the 65 to 79 age group presenting 84% negative SLNB (OR, 2.18) and the over 80 age group with 81% negative SLNB (OR, 2.07). Other factors found to present a significantly higher negative SLNB rates are triple negative breast tumors (OR = 1.83; 83% negative SLNB) and Asian ethnicity (OR = 1.62, 81% negative SLNB).

Conclusion: Our results show breast cancer subpopulations with significantly decreased risk of axillary involvement, for which the use of SLNB could be safely omitted under the appropriate circumstances. This may have implications for surgical decision-making in cases of micro-invasion, and small invasive breast cancers found following excisional biopsy, especially in elderly patients.