70.08 Neoadjuvant Radiation Does Not Increase Perioperative Morbidity in Gastrectomy for Gastric Cancer

Posted on December 22, 2014

Z. Sun1, D. P. Nussbaum1, P. J. Speicher1, B. Czito1, D. Tyler1, D. G. Blazer1  1Duke University Medical Center,Durham, NC, USA

Introduction:
Neoadjuvant radiotherapy (RT) as a component of the multimodality approach to gastric cancer has demonstrated promising results in terms of locoregional recurrence, pathologic response, and survival in several institutional series. However, concerns remain about its effect on perioperative morbidity and mortality.

Methods:
The 2005-2011 National Surgical Quality Improvement Program (NSQIP) Participant Data Use File (PUF) was queried for patients undergoing gastrectomy for gastric cancer. Subjects were classified by use of neoadjuvant RT. Perioperative variables and outcomes were compared.  Groups were then propensity matched using a 3:1 nearest-neighbor algorithm and multivariable logistic regression was performed to assess neoadjuvant RT as a predictor of outcomes. 

Results:
A total of 2,772 patients were identified, among whom 55 (2.0%) underwent neoadjuvant RT.  Patients who received neoadjuvant RT were more likely to have received preoperative chemotherapy within 30 days (p<0.001) and steroids (7.3 vs. 1.6%, p=0.007); they were also more likely to have recent weight loss (30.9 vs. 16.2%, p=0.007).   However, after propensity matching, all baseline characteristics between groups were highly similar.  Concomitant major resection was common for all patients (28.1%), and there was no difference between groups.  Total gastrectomy was more common among patients who underwent neoadjuvant RT (70.9 vs. 46.7%, p=0.003), and operative time was considerably longer (290 vs. 236 minutes, p=0.009).  There were no differences in overall complications (23.6 vs. 29.7%, p=0.489) or 30-day mortality (3.6 vs. 3.6%, p=0.999).  There were also no differences in major complications, surgical site infection, or early reoperation.  

Conclusion:
Neoadjuvant RT was not associated with an increase in 30-day mortality or perioperative morbidity after gastrectomy.  Given reasonable morbidity demonstrated in our study and other institutional series, neoadjuvant RT warrants further investigation as part of the multidisciplinary management of resectable gastric cancer
 

70.09 Major Depression in Colorectal Cancer Survivors: a population-based study of 1783 patients.

Posted on December 22, 2014

J. Liang3, N. Fitzgerald2, S. Ahmed1, D. Hiller1, J. Bohl1, C. J. Clark1  1Wake Forest Baptist Health,Department Of Surgical Oncology,Winston Salem, NC, USA 2Wake Forest Baptist Health,Public Health Sciences,Winston Salem, NC, USA 3Wake Forest School Of Medicine,Winston Salem, NC, USA

Introduction:
Colorectal cancer survivorship has improved significant over the last 20 years due to earlier detection through screening programs and better multimodality treatment options. Few studies have evaluated the health-related quality of life (HRQOL) for colorectal cancer patients and the long-term risk of major depressive disorder (MDD) among survivors. The current study hypothesizes that lower socio-economic status (SES) increases the risk of major depression among colorectal cancer survivors. 

Methods:
A population-based cohort study was conducted based on linkage data from the Surveillance, Epidemiology, and End-Result registry and the Medicare Health Outcome Survey (SEER-MHOS).  Patients with pathology confirmed and resected colorectal cancer over 65 years who completed a survey after diagnosis were identified in the SEER-MHOS.  Survey respondents with a VR12 mental component score (MCS) less than 42 and positive depression screen were defined as patients with MDD.  Using univariate and multiple variable analyses, patients with and without MDD were evaluated to identify risk factors for MDD.   

Results:
1783 resected colorectal cancer patients (median age 77, 50.8% female) were identified in the SEER-MHOS dataset.  The majority of patients were white (n=1337, 75%) and Stage II (n = 603, 43%).  Median time from diagnosis to last follow-up was 130 months with median time from diagnosis to survey of 62 months.  In this cohort, 278 (15.6%) had MDD.  On univariate analysis, risks factors for MDD included female sex, non-white race, less than high school education, income less than $30,000 per year, no home ownership, more than two comorbidities, and inability to perform more than two out of six activities of daily living (ADLs) (all p < 0.05).  Larger tumor size and advanced cancer stage did not predict MDD.  After adjusting for comorbidities, ADLs, and time from diagnosis to survey, income less than $30,000 per year (OR 1.5, 1.02-2.22 95% CI, p = 0.042) and non-white race (OR 1.51, 1.05-2.17 95% CI, p =0.0259) were independent predictors of MDD. 

Conclusion:
In the current study, lower socioeconomic status, specifically patient income, is associated with major depressive disorder among colorectal cancer survivors.
 

70.10 Bleeding Complications Post-Mastectomy for Breast Cancer: Incidence, Predictors, and Economic Impact

Posted on December 22, 2014

I. Nwaogu1, M. Olsen1, J. A. Margenthaler1  1Washington University,Surgery,St. Louis, MO, USA

Introduction: Bleeding complications are serious complications following breast cancer surgery but very few single-institution studies have estimated and reported the incidence. Additionally, no studies have examined the economic impact of postoperative bleeding. The present study aims to determine the incidence of bleeding events following mastectomy, identify potential risk factors for bleeding, and evaluate the impact on hospital utilization and treatment costs.

Methods: Using the 2011 Healthcare Cost and Utilization Project National Inpatient Sample (HCUP NIS), mastectomy hospital discharges for a primary diagnosis of breast cancer were extracted using ICD-9-CM procedure codes 85.34-85.48 for mastectomy and diagnosis codes 174.0-174.9 for breast cancer. Discharges with postoperative bleeding were identified based on ICD-9-CM diagnosis codes 998.11 and 998.12. Discharges with male gender or a history of coagulation disorders were excluded from the study. Overall bleeding complication rates and re-operation rates (for bleeding) were assessed. Risk factors and the impact of bleeding on length of stay and hospital costs were determined using regression analysis. The results were projected to the national level.

Results:7907 discharges coded for mastectomy met inclusion criteria. Of these, 201 had bleeding complications (2.54%), with 42 cases requiring re-operation during the mastectomy admission. On univariate analysis, the presence of congestive heart failure (CHF), obesity, diabetes, chronic pulmonary disease, and the absence of concomitant reconstructive surgery were associated with increased bleeding event. On multivariate analysis, only the presence of CHF remained as a statistically significant predictor of bleeding complications (Odds ratio [95% CI]: 2.45, [1.25, 4.92], p = 0.009). On average, bleeding complications extended the length of stay by 1.3 days (p < 0.0001) while increasing hospital costs by $5,495 per admission (p <0.0001). Projected to the national level, bleeding complications accounted for an additional 1,254 days of hospital stay with costs exceeding $5.3 million.

Conclusion:Bleeding complications following mastectomy for breast cancer had an incidence of 2.5%, with CHF the only identified risk factor for bleeding complications. Finally, such bleeding events, although infrequent, are associated with substantial economic costs in terms of extended length of stay and increased treatment costs.

 

70.11 Placement of a Feeding Jejunostomy Tube Is Safe at the Time of Gastrectomy for Gastric Cancer

Posted on December 22, 2014

D. P. Nussbaum1, Z. Sun1, B. C. Gulack1, J. E. Keenan1, D. S. Tyler1, P. J. Speicher1, D. G. Blazer1  1Duke University Medical Center,Department Of Surgery,Durham, NC, USA 2Duke University Medical Center,Department Of Surgery,Durham, NC, USA

Introduction: Feeding tube placement is common among patients undergoing gastrectomy, and national guidelines currently recommend consideration of a feeding jejunostomy tube (FJT) for all patients undergoing resection of gastric cancer.  However, to date there is no comparative effectiveness data regarding the safety of FJT placement at the time of gastrectomy for gastric cancer.  

Methods: The 2005-2011 ACS-NSQIP Participant User Files were queried to identify patients who underwent gastrectomy for gastric cancer.  Subjects were classified by the concomitant placement of a FJT.  Groups were then propensity matched using a 1:1 nearest neighbor algorithm, and outcomes were compared between groups.  The primary outcomes of interest were overall 30-day morbidity and mortality.  Secondary endpoints included major complications, surgical site infection, and early reoperation.  

Results: In total, 2,980 subjects underwent gastrectomy for gastric cancer, among whom 715 (24%) also had a FJT placed.  Patients who had a FJT placed were more likely to be male (61.6 vs. 56.6%, p=0.02), have recent weight loss (21.0 vs. 14.8%, p<0.01), and to have undergone recent chemotherapy (7.9 vs. 4.2%, p<0.01) and radiation therapy (4.2 vs. 1.3%, p<0.01).  They were also more likely to have undergone total (compared to partial) gastrectomy (66.6 vs. 28.6%, p<0.01) and have concomitant resection of an adjacent organ (40.4 vs. 24.1%, p<0.01).  After adjustment with propensity matching, however, all baseline characteristics and treatment variables were highly similar.  Among propensity-matched groups, there were no statistically significant differences in either primary outcome: 30-day overall complications (38.8 vs. 36.1%, p=0.32) or mortality (5.8 vs. 3.7%, p=0.08).  There were also no differences in major complications, surgical site infection, or early reoperation.  Operative time was slightly longer among patients with feeding tubes placed (median 248 vs. 234 minutes, p=0.01), but otherwise there were no significant differences in any outcomes between groups. 

Conclusion: Placement of a FJT may result in slightly increased operative times, but does not appear to result in increased perioperative morbidity or mortality.  Further investigation is needed to define the patients most likely to benefit from FJT placement.  

70.13 How Effective is the “Two-Week Wait” Target in Improving Survival in Colorectal and Breast Cancer?

Posted on December 22, 2014

K. M. Guest1, I. Nikolopoulos1, M. Kumari1, K. Thakur1  1Queen Elizabeth Hospital,General Surgery,London, LONDON, United Kingdom

Introduction: Incidence rates of colorectal cancer have risen very slowly for two decades, while mortality rates have fallen by over 25%. Incidence rates for breast cancer have continued to rise in the last 20 years chiefly among more affluent women, while mortality rates have fallen by one quarter. Suspected cancer waiting time targets were introduced in the UK to monitor service performance via process improvement. The intention was to improve the outcome (survival) of the disease. The “Two-Week Wait” system denotes that any patient with suspected cancer must be seen by a specialist within 2 weeks of referral. The aim of this study was to assess whether the Two-Week Wait target can improve survival in patients with colorectal and breast cancer.

Methods: 613 and 625 patients were diagnosed with colorectal and breast cancer respectively between January 2002 and December 2006. Data were retrospectively collected from the cancer database at Queen Elizabeth Hospital, England UK. Survival was compared in patients that were referred via the Two-Week Wait system (Group1) and those not referred via this pathway (Group2).

Results: Only 29% of colorectal cancer patients were referred under the Two-Week Wait rule, with the remainder a significant proportion coming from Accident & Emergency (n=131) and General Practitioners (n=144). The mean survival for Group 1 was 3.88 years (1415 days, 95% confidence interval 1283-1546 days) and for Group 2 was 3.41 years (1245 days, 95% confidence interval 1155-1331 days) (p-value=0.005). Kaplan-Meier comparison showed 2-year survival to be 71% for Group 1 and approximately 60% for Group 2.

44% of breast cancer patients were referred under the Two-Week Wait rule and of the remainder a significant proportion came from breast cancer screening (n=122) and GPs (n=135). The mean survival for Group 1 was 4.63 years (1690 days, 95% confidence interval 1604-1775 days) and for Group 2 was 5 years (1825 days, 95% confidence interval 1759-1891 days) (p-value=0.142). Kaplan-Meier comparison showed 2-year survival to be similar in both groups at approximately 83%.

Conclusions: The Two-Week Wait rule significantly improves the survival of patients with colorectal cancer. However the under-utilisation of this referral route cannot be ignored, as an unacceptably high percentage of colorectal cancer patients (30%) come via Accident & Emergency. On the other hand, the Two-Week Wait rule does not appear to affect the survival of patients diagnosed with breast cancer. A well established breast cancer screening programme can lead to equally good survival rates for those referred via the non-Two-Week wait route.

 

70.14 Age Impacts Efficacy of Radiotherapy in Patients with Soft Tissue Sarcoma

Posted on December 22, 2014

N. K. Yuen1, C. Li1, R. Bold1, A. Monjazeb1, D. Borys1, R. Canter1  1University Of California – Davis,Surgical Oncology,Sacramento, CA, USA

Introduction:

Radiation therapy (RT) is a standard component in the multimodality management of soft tissue sarcoma (STS), although increasing studies are focusing on the selective implementation of this modality. We hypothesized that the effects of RT would be influenced by age, potentially contributing to differences in treatment outcome. We sought to evaluate the effect of age on RT in a national database of STS patients.

 

Methods:

Using the Surveillance Epidemiology and End Results registry (1990-2011), we identified 30,898 adult patients (>18) with non-metastatic STS of major histologies undergoing surgery with or without RT. We compared patient demographics, tumor characteristics, and treatments by age. Multivariable analyses were used to examine the effect of these variables on overall (OS) and disease-specific survival (DSS). Hazard ratios were calculated based on multivariable Cox proportional hazards models.

 

Results:

The mean age at diagnosis was 54.9 years old. 31.5% of patients were ≥ 65 years old. 47.5% of patients were male, and 66.6% were white. Histologic grade was 32.4% high, 12.9% intermediate, 13.5% low, and 41.3% unknown. Leimoyosarcoma was the most common histology (27.9%), and extremity was the most common tumor site (37.4%). 70.2% of patients underwent surgery alone, 25.1% received adjuvant RT, and 4.7% neoadjuvant RT. On multivariable analysis, gender, year of diagnosis, histology, grade, size, marital status, and RT modality predicted OS, while year of diagnosis, ethnicity, histology, site, grade, RT, size, and marital status predicted DSS. The hazard ratio (HR) for OS was 1.11 (95% CI 1.02-1.21) and DSS was 1.13 (95% CI 1.03-1.25) comparing neoadjuvant to adjuvant RT in all age groups. In patients < 65 years old, the HR for OS was 1.23 (95% CI 1.09-1.39) and DSS was 1.24 (95% CI 1.10-1.39) for neoadjuvant to adjuvant RT. For patients ≥ 65, the HR for OS was 0.98 (95% CI 0.85-1.13), and DSS was 0.99 (95% CI 0.83-1.17) for neoadjuvant to adjuvant RT.

 

Conclusions:

Adjuvant RT was associated with superior oncologic outcome compared to neoadjuvant RT among STS patients undergoing surgical resection. The negative effect of neoadjuvant RT was exaggerated in younger patients, suggesting an age-related effect of RT. Further investigation into the etiology of this age-related association is indicated.

 

70.15 Defining the Burden of Rectal Cancer: A Population Study

Posted on December 22, 2014

D. Hayden1, M. Ostrowski2, T. Markossian3, J. Eberhardt1, T. Saclarides1  1Loyola University Medical Center,Department of General Surgery,Maywood, Illinois, USA 2Loyola University Medical Center,Stritch School Of Medicine,Maywood, Illinois, USA 3Loyola University Medical Center,Department Of Public Health Sciences,Maywood, Illinois, USA

Introduction:  The treatment course of rectal cancer is long and wrought with morbidity.  The burden of this disease is underappreciated and impacts patients, their families and the health care system.

Methods:  Secondary analysis of the 2011 National Inpatient Sample was performed using the ICD-9CM codes for rectal cancer, most common secondary diagnoses and co-morbidities. 

Results: For the 24,575 rectal cancer discharges in 2011, mean age was 64.2 (17-99); 58.1% were male.  69.8% were white, 9.3% black, 7.7% Hispanic and 2.8% Asian/Pacific Islander.  5.6% were obese and 3.1% morbidly obese.  Most common comorbidity listed as the secondary diagnosis was hypertension (4.6%).  50.6% of patients resided in or near metropolitan areas with population ≥ 1 million; 16.5% from those 250,000-999,999, 9.7% from those 50,000-249,999 and 11.3% from micropolitan areas.  Inpatient admissions seemed to be evenly distributed across income quartiles. Medicare (46.5%) and private insurance (36.8%) were most common; Medicaid accounted for 9.3% of payer distribution and 3.8% was self-pay.  Admissions were most frequently elective (72.4%).  Mean length of stay was 7.42 (0-277) days.  47.8% of patients were discharged home; 35.4% required home health nursing; 13.9% were transferred to rehab or skilled nursing. 2.4% of patients died during the hospital admission.  For the obese and morbidly obese rectal cancer patients, LOS was increased: 8.27 days (0-64), were younger 60.4 (27-89), higher percentage were white (74.0%) and more had private insurance (48.6%, all p=0.00).    

Conclusion: Rectal cancer impacts patients and the health care system due to lengthy hospitalizations and postoperative healthcare utilization of home health nursing and short-term facilities. Obesity and related comorbidities contribute to poorer outcomes; however, lifestyle interventions at diagnosis may decrease this burden.

 

70.01 Ventriculoperitoneal shunt in patients with CNS neoplasms: An analysis of 59 cases

Posted on December 22, 2014

F. Nigim1, J. Critchlow1, E. Kasper1  1Beth Israel Deaconess Medical Center,Department Of Surgery,Brookline, MA, USA

Introduction:
About 1-5% of patients with cerebral metastasis suffer from hydrocephalus, and so do about 40% of patients with primary brain tumors. These patients often carry a poor prognosis. The aim of the present study is to reassess the validity of VPS placement with the assistance of the general surgeon in the abdominal part in oncological patients.

Methods:
 A total of 59 patients underwent first time VPS placement at BIDMC between 2004 and 2012; 40 had hydrocephalus from brain metastasis and 19 from primary tumors. Analyzed independent variables included demographics, BMI, PMH, clinical presentation, indication for surgery, Karnofsky performance status (KPS) score, and surgical technique; dependent variables were post-operative symptoms and occurrence, cause, and time of shunt failures. Outcomes were analyzed via T-test and Kaplan-Meier estimates for shunt survival.

Results:
Mean age was 57.2 years; and mean operative time was 50.4 minutes. Symptomatic palliation was achieved in 93%; patients with severe symptoms (e.g. debilitating headaches and nausea and vomiting) did benefit significantly from VPS placement. Mean follow-up time was 6.3 months, complications occurred in only 6.7% (n=4) patients during follow up; with 2 wound infections treated with antibiotics, and 2 shunt obstructions requiring revision. Initial KPS and 3-month KPS were 65 ± 16.4 and 75 ± 16.0, respectively. We started with 59 patients, 16 (27.1%) patients were alive at 6-month and 10 (16.9%) patients were alive at 1-year, in all survived patients the shunt was functioning. 

Conclusion:
VPS remains a valid option for cancer patients with low KPS and improves the quality of life in those patients, even in the setting of previous infection, hemorrhage, or leptomeningeal disease since shunt patency outlasts the overall survival of nearly all patients. 
 

69.15 Clinicopathologic Factors Associated With a False Negative AUS in Patients With Breast Cancer

Posted on December 22, 2014

I. Nwaogu1, Y. Yan1, J. A. Margenthaler1  1Washington University,Surgery,St. Louis, MO, USA

Introduction: Axillary ultrasound (AUS) has been used in an attempt to identify sub-clinical node-positive disease and improve clinical staging and treatment recommendations. We sought to identify clinicopathologic factors potentially related to false negative AUS results.

Methods: Patients with a clinically node-negative Stage I-II invasive breast cancer who also had a normal AUS were identified from our prospectively maintained database. All AUS studies were performed by dedicated breast radiologists and interpreted as “normal” according to the absence of specific characteristics previously shown to be more commonly associated with metastatic involvement. True- and false-negative AUS studies were compared statistically based on clinical, radiographic, and histologic parameters.

Results:Of the 118 patients with a normal AUS, 25 (21%) were ultimately found to be node-positive on final pathologic assessment following axillary surgery. On bivariate analysis, primary tumor size, lymphovascular invasion (LVI), and Her2neu receptor status were found to be significantly different between true- and false-negative AUS. The average tumor size was smaller in the true-negative group compared to the false-negative group [16 vs 21mm (p< 0.01)]. The presence of LVI was more likely in the false-negative group [11/25 (44%)] compared to the true-negative group [7/93 (8%)] (p< 0.0001). Her2neu receptor status was more likely amplified in the false-negative group [8/25 (32%)] compared to the true-negative group [13/93 (14%)] (p=0.037). No significant difference was noted between groups with regard to patient age, race, body mass index, tumor grade, histologic type, estrogen or progesterone receptor status, and time between AUS and axillary surgery. On multivariate analysis, only the presence of LVI achieved statistical significance (p=0.0007).

Conclusion:AUS is a valuable tool that accurately predicted absence of axillary disease in 79% of patients with clinically node-negative breast cancer. AUS findings may be less accurate in the setting of LVI, and a negative AUS in patients with this feature should be interpreted with caution.

 

69.16 Use of a Recurrence Score In Locally Recurrent/New Primary Breast Cancer

Posted on December 22, 2014

N. C. Vera4, D. Carr4, J. Mullinax1, D. Korz1, W. Sun1, C. Laronga1, S. Hoover1, W. Fulp2, G. Acs3, M. C. Lee1  1Moffitt Cancer Center And Research Institute,Breast Program,Tampa, FL, USA 2Moffitt Cancer Center And Research Institute,Biostatistics,Tampa, FL, USA 3Women’s Pathology Consultants,Ruffolo Hooper & Associates,Tampa, FL, USA 4University Of South Florida College Of Medicine,Tampa, FL, USA

Introduction: Treatment planning for locally recurrent estrogen receptor positive (ER+) invasive breast cancer is controversial. The Recurrence Score (RS) from the 21-gene breast cancer assay (ODX) is commonly used for primary early stage ER+ invasive breast cancer for adjuvant treatment recommendations, but not generally obtained in locally recurrent ER+ tumors. We reviewed our experience with RS first performed on the recurrent tumor.  

Methods: An IRB-approved, single-institution retrospective chart review of a prospective ODX database was performed. Most patients had ODX performed on an initial invasive breast cancer (1o cancer); another set of patients, the minority, had ODX performed on an ipsilateral local recurrence/new primary (2nd cancer); none had clinical evidence of concurrent regional or distant metastasis at presentation. Performance of the ODX assay was based on NCCN guidelines (1o cancer) or physician discretion (2nd cancer). Data collected included demographics, clinical-pathologic variables, surgery type, RS, adjuvant treatment and outcomes. Comparisons between patients with 1o breast cancer and patients with 2nd cancers were made by general linear regression model and the exact Wilcoxon Rank Sum Test. 

Results:594 patients with 1o breast cancer and 7 patients with 2nd breast cancer had ODX and RS.  Median age of patients at time of ODX was 58 years (range 27-84) for 1o cancer and 58.5 years (range 36-63) for 2nd cancers (p=0.411) respectively. The majority of patients with a 2nd breast cancer had a prior history of breast conservation (6/7). Median invasive tumor size of 1o cancer was 1.5cm and 2nd cancer was 1.4cm. One 2nd cancer was ILC, otherwise all documented 1o and 2nd breast cancers were ER+ and of invasive ductal histology. For 2nd cancers, median time from 1o breast cancer surgery to diagnosis with 2nd breast cancer was 92 months (range 13-120)[Table of 2nd cancers]. Median RS was higher in patients with 2nd cancer at 22 (range 15-37) compared to 1o cancer at 16 (range 0-63) (p=0.03). Categorically, more 2nd cancer patients had a high RS (28.6%) than those with 1o cancer (8.1%) but it was not significant (p=0.08). Tumor size, nodal status, degree of ER expression, nuclear grade, number of mitoses, and compliance with endocrine therapy were not significantly different between patients with 2nd cancer and 1o cancer. Overall survival tended to be better in patients with 1o breast cancer (P=0.049).

Conclusion:Performance of ODX in ER+, locally recurrent/new primary invasive breast cancer (2nd breast cancer) should be considered for prognostication and adjuvant systemic treatment recommendations.
 

69.17 Menopausal Status Affects Presentation but Not Outcome in Invasive Lobular Carcinoma

Posted on December 22, 2014

M. Zamanian1, A. Soran1, M. K. Wright1, C. Thomas1, G. M. Ahrendt1, M. Bonaventura1, E. J. Diego1, R. R. Johnson1, P. F. McAuliffe1, K. McGuire1  1University Of Pittsburgh School Of Medicine,Division Of Surgical Oncology, Department Of Surgery,Pittsburgh, PA, USA 2Magee Women’s Hospital Of UPMC,Surgical Oncology,Pittsburgh, PA, USA

Introduction: Invasive lobular carcinoma (ILC) is the second most common form of breast cancer, with rates increasing over the past 10-15 years. ILC has unique molecular and clinical characteristics, distinct from invasive ductal carcinoma. Studies of ILC, especially in the premenopausal population, remain limited. We hypothesize that premenopausal patients diagnosed with ILC will present with higher stage tumors and with lower estrogen receptor (ER) expression and will have poorer disease free and overall survival (DFS, OS).

Methods: A retrospective review of a prospectively collected database identified all pre- and postmenopausal patients treated for ILC at a single institution between 2004 and 2010. Patient and tumor characteristics were collected as well as outcome data. Patients whose menopausal status at diagnosis was not recorded were categorized as postmenopausal if they were over age 50 and premenopausal if they were under age 45. Patients age 46-49 were considered perimenopausal and were excluded. ER/PR (progesterone receptor) expression was measured using H-score (H-score = sum of % nuclear staining x intensity 0, 1+, 2+, 3+, giving a range from 0-300).  The two groups were compared for differences in presentation and outcome using uni- and multivariate analysis.

Results: 87 premenopausal and 226 postmenopausal patients were treated for ILC during the study period. Demographics were well balanced between the groups with the exception of age (p<0.001). Univariate analysis showed significantly larger tumor size (3.4±2.9 versus 2.5±2 cm, p=0.002), higher clinical stage (p=0.003), higher PR H-score (158±91 versus 121±100,p=0.005), and lower ER H-score (216±67 versus 242±65, p=0.004) in premenopausal versus postmenopausal patients, respectively. Significant differences in treatment with surgical and systemic therapy were also identified (Table 1). HER2 status was similar between groups. Multivariate analysis showed menopausal status to be independently predictive of tumor size (p=0.012), ER and PR (p<0.001) H-score and likelihood of receiving chemotherapy (p=0.013). DFS and OS were similar between the two groups (p=0.14 and p=0.16, respectively).

Conclusion: In this retrospective review of patients with ILC, premenopausal patients presented with larger tumors and with lower ER H-scores than post-menopausal patients. Despite these adverse clinical factors, there were no significant differences in DFS or OS. Premenopausal patients received systemic therapy more often, which may have contributed to equivalent outcomes. Further research is needed to understand how local and systemic therapy affect outcome in premenopausal patients with ILC.

69.18 Breast Cancer Outcomes in a Population with a High Prevalence of Obesity

Posted on December 22, 2014

V. C. Herlevic1, R. S. Mowad1, J. K. Miller1, N. A. Darensburg1, B. D. Li1, R. H. Kim1  1Louisiana State University Health Sciences Center – Shreveport,Surgery,Shreveport, LA, USA

Introduction:

Obesity has been associated with poor prognosis in breast cancer. However, most previous studies examined populations with relatively low proportions of obese patients. Given that forecasts predict obesity rates to exceed 50% by 2030, it is important to examine breast cancer outcomes in populations with higher rates of obesity. We hypothesized that obesity, as measured by Body Mass Index (BMI), is associated with decreased overall survival and disease-free survival in patients with invasive breast cancer in a population with a high prevalence of obesity.

Methods:

A retrospective review of a prospectively maintained database was conducted on patients treated for invasive breast cancer at an academic medical center between July 1987 and May 2013. BMI was calculated from each patient’s height and weight at time of diagnosis. Patients were categorized as normal (BMI < 25 kg/m2), overweight (BMI 25 – 30 kg/m2), or obese (BMI > 30 kg/m2), as per the definitions established by the World Health Organization. The endpoints of overall survival and disease-free survival were analyzed.

Results:

A total of 740 patients with invasive breast cancer were included for analysis. Based on BMI, 127 (17.2%) were categorized as normal, 203 (27.4%) were overweight, and 410 (55.4%) were obese. The median follow-up was 49 months. There were 17 deaths (13.3%) in normal patients, 34 (16.7 %) in overweight patients, and 64 (15.6%) in obese patients (p=0.74). By Kaplan-Meier survival analysis, there were no differences in overall survival (p=0.91) or in disease-free survival (p=0.99) between the three groups.

Conclusion:

Obesity is not associated with decreased overall or disease-free survival in a patient population with a high prevalence of obesity. These findings suggest that there may be other factors that contribute to the poor prognosis of obese breast cancer patients observed in populations with lower rates of obesity.

69.19 Breast Density, BMI, and Outcomes in Premenopausal Women with Breast Cancer

Posted on December 22, 2014

M. K. Wright1, A. Soran1,2, M. Zamanian1, C. Thomas2, G. M. Ahrendt1,2, M. Bonaventura1,2, E. J. Diego1,2, R. R. Johnson1,2, P. F. McAuliffe1,2, K. P. McGuire1,2  1University Of Pittsburg,School Of Medicine,Pittsburgh, PA, USA 2Magee Women’s Hospital Of UPMC,Surgical Oncology,Pittsburgh, PA, USA

Introduction:  

Breast density is a well-established risk factor for the development of breast cancer.  Some studies suggest that breast density is associated with a tumorigenic microenvironment within the breast, leading to more cancers and higher stage at presentation. We hypothesize that premenopausal women with dense breasts will present with breast cancer at a  higher stage and will have worse long-term outcomes compared to their non-dense breasted counterparts. 

Methods:

We performed a retrospective study of a prospectively collected database identifying premenopausal women with breast cancer who presented to a single institution between 2006 and 2010.  Premenopausal status was defined as age less than 50 years or no menstrual period for at least one year. Patient and tumor characteristics were collected, as well as long-term outcomes.  Patients were stratified into two categories of breast density: ‘non-dense’, defined as fatty replaced  or scattered fibroglandular densities (BIRADS Density Categories 1 & 2, respectively) and ‘dense’, defined as heterogeneously dense or homogeneously dense (BIRADS Density Categories 3 & 4, respectively). The two breast density groups were compared for differences in presentation and outcomes using univariate and multivariate analyses.

Results:

477 premenopausal women with breast cancer were identified.  Clinicopathologic factors, including age, race, tumor histology, receptor status, and treatment types were well balanced between the two breast density groups.  On univariate analysis, breast density did not correlate with stage at presentation, tumor grade, lymphovascular invasion, clinical stage, treatment type, surgery type or overall survival.  Lower breast density was strongly correlated with higher BMI, poorer disease free survival (DFS), and larger tumor size on univariate analysis (Table 1).  Multivariate analysis also showed that BMI (p = 0.05) and tumor size (p = 0.001) were significantly associated with DFS, whereas breast density was not.  

Conclusion: 

In this retrospective study of premenopausal women with breast cancer, we found that higher breast density was not associated with higher stage at presentation or with poorer outcomes.  This study suggests that higher BMI and tumor size at presentation, although related to breast density, are more predictive of recurrence than breast density in premenopausal patients. Further study is needed to elucidate the link between BMI, breast density and outcome in premenopausal breast cancer.
 

69.20 Is routine excision of Pseudoangiomatous Stromal Hyperplasia (PASH) an unnecessary surgery?

Posted on December 22, 2014

D. R. Layon1, A. D. Brooks2  1Drexel University College Of Medicine,Philadelphia, Pa, USA 2University Of Pennsylvania,Endocrine And Oncologic Surgery,Philadelphia, PA, USA

Introduction:  Pseudoangiomatous stromal hyperplasia (PASH) is a benign breast lesion of mesenchymal origin for which the optimal treatment strategy is unclear. We performed a systematic review of the literature in order to determine if excision is necessary for managing PASH found on a biopsy. 

Methods:  Keyword searches for “pseudoangiomatous stromal hyperplasia” and “pseudoangiomatous hyperplasia of the mammary stroma” were queried in PubMed. Exclusion criteria were: articles not about relevant pathology, extra-mammary lesions, and inability to extract PASH data from mixed patient cohorts. Eligible articles were reviewed for patient demographics and study characteristics. We determined whether each reported cancer was incidentally or directly related to the PASH lesion. Incidental malignancies were: detected independently, temporally or spatially separate from the PASH lesions, or contained only microscopic foci of PASH. Malignancies were classified as directly related if located within PASH lesions. 

Results: The search returned 122 results; 6 studies were obtained from other sources for a total of 128 studies. 18 studies were excluded. Eligible articles included: 51 case studies, 41 case series 17 review articles, and 1 textbook. The articles had a population of 1,508 individuals. 104 were male, 1,394 were female and 10 were not reported. 73% of PASH cases were managed via excision, 10% by observation, 1% by mastectomy, 0.79% by other methods (incisional biopsy, mammoplasty, or removal via core needle or vacuum-assisted biopsy), and 14% of cases had no reported management. There were 91 cases of malignant or pre-malignant lesions reported in the PASH literature (Table 1): invasive or noninvasive lesions, subtype not specified (48 cases), invasive ductal carcinoma (14), ductal carcinoma in-situ (DCIS, 16), Non-Hodgkins Lymphoma (3), invasive micro papillary carcinoma (1), invasive adenocarcinoma (5), and myofibroblastic sarcoma (4). Of the 91 cases, 34 were temporally separate, 21 spatially separate, 22 were identified independently of PASH, 8 had only microscopic foci of PASH and 5 were unable to be classified. The other 5 malignancies appear to be four cases of primary myofibroblastic sarcomas arising from PASH tumors and a single case of DCIS found within a PASH tumor. Thus, we estimate the incidence of malignant transformation of PASH lesions at 5/1508 or 0.3%. 

Conclusion: Although most PASH lesions in the literature were treated by excision, the reported incidence of malignant transformation of PASH lesions is 0.3%. Based on the low incidence of malignancies arising from PASH lesions we conclude that the diagnosis of PASH alone, in the absence of other indications, does not require surgical intervention. 

69.03 Delaying I-131 Treatment For Papillary Thyroid Cancer Is Not Associated With Increased Recurrence

Posted on December 22, 2014

P. H. Dedhia1, S. Grzegorski1, M. S. Cohen1, B. S. Miller1, P. G. Gauger1, D. T. Hughes1  1University Of Michigan,Surgery,Ann Arbor, MI, USA

Introduction:  Radioactive iodine (RAI) ablation is frequently used in patients with papillary thyroid cancer (PTC) larger than 1 cm to ablate the thyroid remnant and facilitate detection of recurrent disease. Treatment with RAI is usually performed 6-8 weeks after thyroidectomy. Here we assess the effect of delay in RAI on post-RAI thyroglobulin levels and PTC recurrence. 

Methods:  This is a retrospective review of 447 patients with pathologically confirmed PTC from thyroidectomy procedures performed between 2009 to 2013 in the endocrine surgery division at the University of Michigan. Clinical characteristics, pathology, postoperative stimulated thyroglobulin (Tg) levels, days elapsed from surgery to radioactive iodine treatment, unstimulated Tg levels after RAI, and recurrence were examined.

Results: Of the 447 patients with papillary thyroid cancer from 2009 to 2013 who underwent thyroidectomy, 50.5% (n=226) were treated with RAI. Of the patients treated with RAI, 37% of patients underwent RAI greater than 60 days after thyroidectomy. The average tumor size in patients treated with immediate RAI and patients who underwent delayed RAI were 2.2 cm and 2.1 cm, respectively. 89% of patients with immediate RAI had undetectable Tg compared to 83% of patients with delayed RAI (p=0.20). Of patients that underwent immediate RAI, 5.9% (n=8) developed recurrence, whereas 2.5% of patients undergoing delayed RAI developed recurrence (n=2; p = 0.25). 

Conclusion: Administration of RAI in a delayed manner is not associated with decreased likelihood to achieve undetectable Tg or with increased recurrence of PTC after RAI treatment. 

 

69.04 Body Mass Index and Vitamin D Deficiency in Multi-gland Primary Hyperparathyroidism

Posted on December 22, 2014

J. A. Glenn1, T. W. Yen1, D. B. Evans1, T. S. Wang1  1Medical College Of Wisconsin,Division Of Surgical Oncology – Section Of Endocrine Surgery,Milwaukee, WI, USA

Introduction: Sporadic primary hyperparathyroidism (pHPT) is due to multi-gland disease (MGD) in up to 15% of cases. MGD may occur due to systemic disease and/or variations in the parathyroid hormone (PTH)-calcium-vitamin D axis. Vitamin D deficiency is more pronounced in obese patients due to sequestration of fat-soluble vitamins in adipose tissue. The aim of this study is to explore the relationship between obesity, vitamin D deficiency, and MGD.

Methods: This is a retrospective chart review of a prospective parathyroid database of 583 patients with sporadic pHPT who underwent parathyroidectomy between 1/1/2009 – 12/31/2013. Demographic, clinicopathological (number and weight of enlarged glands), and laboratory (serum calcium, PTH, 25-OH vitamin D, creatinine) data were collected.  Patients were equally divided into 4 groups based on body mass index (BMI; kg/m2). The association between BMI and age, gender, race, perioperative vitamin D, PTH, and calcium levels were assessed. Multivariable logistic regression models were performed to determine clinicopathological predictors of MGD.

Results: Mean age of the cohort was 59.8 ± 12.0 years (range, 18-90); 463 (79%) were women, and 507 (87%) were Caucasian. Of these, 103 (18%) patients had MGD. Patients with a higher BMI tended to have higher preoperative PTH levels, although this was not statistically significant (Table 1). There was an association between BMI and vitamin D deficiency (<30 ng/mL, p<0.001). In unadjusted analyses, there was no association between BMI (Table 1), age, gender, or race and the occurrence of MGD. Of the 549 patients with recorded gland weights, the median gland weight was 470 mg (interquartile range, 240-923). There was no association between BMI and gland weight or vitamin D deficiency and gland weight in patients with MGD. On multivariate logistic regression, after adjusting for preoperative calcium and PTH levels, BMI and vitamin D levels were not predictors of MGD.

Conclusion:  There is no association between obesity, vitamin D deficiency and occurrence of MGD in pHPT patients, in adjusted analyses. In addition, BMI and vitamin D deficiency are not associated with increased gland weight for MGD patients. Further studies evaluating the molecular influences of obesity and vitamin D deficiency on parathyroid tissue are needed to help clarify relationships between these factors and MGD.

 

69.05 The Underappreciated Problem of Thyroid Cancer and Hypothyroidism

Posted on December 22, 2014

S. Zaheer1, L. E. Kuo1, J. C. Morrison1, H. Wachtel1, G. C. Karakousis1, D. L. Fraker1, R. R. Kelz1  1University Of Pennsylvania,Department Of Endocrine And Oncologic Surgery,Philadelphia, PA, USA

Introduction:  

Hypothyroidism has been estimated to affect 3.7% of the United States population.  It has been suggested that hypothyroidism may be a risk factor for thyroid cancer.  We sought to examine the association between hypothyroidism and thyroid cancer characteristics.

Methods:

We identified thyroid surgery patients enrolled in our institutional prospective endocrine surgery registry from 2007 to 2013. Patients were included in the study if they were either hypothyroid (Thyroid stimulating hormone (TSH) >4.0 mIU/L) or euthyroid (TSH between 0.4-4.0 mIU/L) at the time of surgical intervention. The primary outcome of interest was thyroid cancer.  Tumor characteristics were examined in the subset of patients with thyroid cancer identified on pathologic review.   Univariate analyses were performed using the Wilcoxon ranksum test, Fisher’s exact test and Chi square test, as appropriate.

Results:

Of 901 patients included in the study, 537 patients were euthyroid and 364 patients were hypothyroid. There were 634 patients (70%) with a confirmed diagnosis of thyroid cancer. Age, gender and race were evenly distributed in the two groups. Patients with hypothyroidism had an increased incidence of cancer compared with euthyroid patients (88% vs. 58% respectively, p<0.01). Tumor characteristics were more aggressive among hypothyroid patients (see table).

Conclusion:

Amongst a population of thyroid surgery patients, there was a significant association between thyroid cancer and elevated TSH.  Moreover, the tumor characteristics associated with hypothyroidism were more aggressive than those seen in the euthyroid state. Clinicians should carefully screen patients with elevated TSH level for nodular disease and, when appropriate, evaluate for thyroid malignancy in a timely fashion.

 

69.06 Increased Rate of Incidental Papillary Thyroid Cancer in Surgical Patients with Benign Thyroid Disease

Posted on December 22, 2014

A. R. Marcadis1, S. Liu1, M. Rodriguez1, J. I. Lew1  1University Of Miami Miller School Of Medicine,Division Of Endocrine Surgery,Miami, FL, USA

Introduction: Patients who undergo surgical resection for benign thyroid disease may have incidental papillary thyroid cancer (PTC) discovered on final pathology. The incidence of PTC has historically been between 5 to 10% in patients treated operatively for benign thyroid disease. This study attempts to determine if there are any preoperative factors for incidental PTC that may help determine the extent of thyroidectomy in patients with benign thyroid disease.

Methods: A retrospective review of prospectively collected data of 1822 consecutive patients who underwent thyroidectomy at a single institution was performed. Of these patients, 355 underwent surgical resection for benign thyroid disease. Indications for surgery were obstructive or compressive symptoms (n=142), hyperthyroid symptoms including Graves’ disease (n=111), goiter size >4 cm (n=92), and substernal goiter (n=10). Patients with indeterminate or malignant preoperative FNA results were excluded. Of all patients, 74% (n=263) underwent total thyroidectomy, and 26% (n=93) underwent a lobectomy. Benign final pathology included nontoxic multinodular goiter (MNG) (n=136), toxic MNG (n=75), nontoxic solitary nodule (n=67), toxic solitary nodule (n=21) and Graves’ disease (n=6). Incidental cancers were defined as PTC discovered only upon final pathology. Age and gender were examined to determine if certain subsets of these categories had a higher indication for incidental cancers, using a two-tailed Z-test at a significance of 0.05.

Results: Overall, 14% (50/355) of patients who underwent surgical resection for benign thyroid disease had incidental PTC on final pathology. Women constituted 90% (n=322) of patients. There was no significant difference between incidental PTC rates in women (14%, 46/322) and men (12%, 4/33). Patients treated for obstructive symptoms had the highest incidental PTC rate at 19% (27/142), followed by goiters >4 cm at 15% (14/92), hyperthyroidism at 8% (9/111). Patients <50 years of age with benign indications for surgical resection had incidental PTC rates of 18% (33/187) compared to a 10% (17/169) incidental PTC rate for patients >50 years of age. Patients <50 years of age with benign thyroid disease had a significantly higher percentage of incidental PTC on final pathology than those patients >50 years of age (p=0.0394).

Conclusion: There is a higher than expected rate of incidental PTC in patients who undergo operations for benign thyroid disease, especially in patients <50 years of age. Therefore, total thyroidectomy by an experienced surgeon should be strongly considered when managing benign thyroid disease in such patients.

 

69.08 Family History of Thyroid Cancer Correlates with More Aggressive Papillary Thyroid Cancer Variants

Posted on December 22, 2014

A. R. Marcadis1, S. Liu1, M. Rodriguez1, J. I. Lew1  1University Of Miami Miller School Of Medicine,Division Of Endocrine Surgery,Miami, FL, USA

Introduction: Thyroid malignancy is the most common endocrine cancer in the United States. Papillary thyroid cancer (PTC) is the most common form, comprising 85% of all thyroid cancers. Although most PTC occurs sporadically, about 5% may be familial in origin. Furthermore, family history has shown to be an influential risk factor for papillary thyroid cancer (PTC). Whether these familial forms of PTC are more aggressive than its sporadic form, however, remains unclear. This study determines if patients with a family history of thyroid cancer harbor more aggressive variants of PTC than patients without a positive family history.

Methods: A retrospective review of prospectively collected data of 1822 consecutive patients who underwent thyroidectomy at a single institution was performed. Patients with malignancies other than PTC were excluded from the study (n=447). Remaining patients were divided into 2 groups: those patients who had a history of PTC in first degree relatives (n=39), and those who did not (n=1336). Patients with PTC on final pathology were further subdivided into those with less aggressive (classic and follicular), and those with more aggressive (diffuse sclerosing and tall cell) variants of PTC. A two tailed Z test at a significance level of 0.05 was used to compare both groups.

Results: Of the 1375 patients included, 3% (39/1375) of patients had a family history of PTC in first degree relatives. Of these patients with a positive family history, 54% (21/39) had PTC on final pathology whereas 28% (374/1336) of patients with no family history had PTC on final pathology. Patients with a family history of PTC had 21% (8/39) follicular, 10% (4/39) diffuse sclerosing, 10% (4/39) classic, 8% (3/39) combined follicular and classic, and 5% (2/39) tall cell variants of PTC.  Patients with no familial history of PTC had 19% (257/1336) follicular, 4% (55/1336) classic, 2% (25/1336) both classic and follicular, 1% (9/1336) both diffuse sclerosing and tall cell, 1% (10/1336) tall cell only, and 1% (18/1336) diffuse sclerosing only variants of PTC. Of those patients with a positive family history, there was a significantly increased incidence of more aggressive variants (diffuse sclerosing and tall cell) of PTC at 15% (6/39) compared to those patients with no family history of PTC at 3% (37/1336) (p<0.05).

Conclusion: Patients with a positive family history of PTC have a significantly higher incidence of more aggressive PTC variants. Any positive PTC family history, therefore, should be considered a risk factor for more aggressive disease that may necessitate earlier recognition and treatment for PTC in these patients and their affected family members.

 

69.09 Surgeon Performed Ultrasound Can Predict More Aggressive Variants of Papillary Thyroid Cancer

Posted on December 22, 2014

A. R. Marcadis1, B. Wang1, M. Rodriguez1, S. Liu1, J. I. Lew1  1University Of Miami,Division Of Endocrine Surgery,Miami, FL, USA

Introduction: Surgeon performed ultrasound (SUS) has been shown to be a valuable imaging modality in the preoperative prediction of malignancy or benignity of thyroid nodules. While patients with papillary thyroid cancer (PTC) are generally considered to have an excellent prognosis, there are some variants of PTC that exhibit more aggressive behavior and may constitute a worse prognosis. The purpose of this study is to determine if SUS can further predict not only malignancy in thyroid nodules, but also more aggressive variants of PTC.

Methods: A retrospective review of 1822 consecutive patients who underwent thyroidectomy at a single institution was performed. Only patients with final pathology confirming the presence of variant PTC and who underwent SUS prior to thyroidectomy were included (n=400). More aggressive variants of PTC were defined as diffuse sclerosing and tall cell variants, whereas classic and follicular variants of PTC were considered less aggressive. SUS characteristics of thyroid nodules were studied, including type (solid, mostly cystic, mixed), calcifications (micro-, coarse, none), borders (irregular, regular), shape (taller greater than wider), and echogenicity (hyper-, hypo-, isoechoic). SUS features were correlated with final histopathology using univariate regression analysis. For SUS features that are statistically significant (p<0.05), odds ratio (OR), confidence interval (CI), and p-values are presented.

Results: On final histopathology, 49 patients (12%) had more aggressive variants of PTC. On univariate analysis, hypoechogenicity (p<0.005, OR=2.700, CI [1.382-5.248]), microcalcifications (p<0.005, OR=2.389, CI [1.306-4.369]) and irregular border (p=0.043, OR=1.853, CI [1.014-3.387]) were SUS features that correlated significantly with more aggressive variants of PTC. Combining 2 of these features increased the predictive value of SUS for aggressive variants (hypoechogenicity and microcalcifications p<0.005, OR=5.757, CI [2.192-15.121]; hypoechogenicity and irregular border p<0.005, OR=3.810, CI [1.590-9.125]; irregular border and microcalcifications p<0.005, OR=2.787, CI [1.384-5.613]). The combination of all three SUS features of hypoechogenicity, microcalcifications, and irregular border had the strongest association for aggressive variants of PTC (p<0.005, OR = 7.054, CI [2.418-20.574]).

Conclusion: The combination of all three SUS features of hypoechogenicity, microcalcifications, and irregular border strongly predicts more aggressive variants of PTC with a 7-fold higher likelihood of being diffuse sclerosing or tall cell variants on final pathology. SUS, therefore, can potentially predict the biologic behavior of PTC at time of diagnosis, and help guide the surgeon in the operative and peri-operative management of these patients.

« Previous Page
Next Page »