13.19 Which Margin-Negative Patients Will Have Occult Disease?: Results from the SHAVE Trial

Posted on January 18, 2016

A. B. Chagpar1, N. Horowitz1, B. Killelea1, T. Tsangaris2, M. Butler1, K. Stavris1, V. Bossuyt1, M. Harigopal1, F. Li1, X. Yao1, S. Evans1, E. Hofstatter1, T. Sanft1, D. Lannin1, M. Abu-Khalaf1, L. Wilson1, L. Pusztai1 2Thomas Jefferson University,Surgery,Philadelphia, PA, USA 1Yale University,New Haven, CT, USA

Introduction: Obtaining negative margins at partial mastectomy (PM) is often seen as being consistent with adequate resection; however, some women with negative margins have occult disease remaining in the breast. We sought to identify factors that are associated with such occult disease.

Methods: The SHAVE trial is a prospective randomized controlled trial in which 235 women with stage 0-3 disease undergoing PM were randomized intraoperatively to either have routine cavity shave margins (CSM) taken at the time of initial surgery (n=119), or not (n=116). The 76 women randomized to the ‘shave’ arm of the trial who had negative margins prior to randomization formed the cohort of interest for this analysis. Patients who were found to have occult cancers in the CSM were compared to those with no further disease. Non-parametric statistical analyses were performed using SPSS Version 21.

Results: Of 76 patients with negative margins prior to randomization to the ‘shave’ group, 9 (11.8%) had occult cancers found in the CSM. Six of these patients had occult DCIS, two had occult invasive disease, and one had both. The median extent of DCIS and invasive cancer found in the CSM was 0.25 cm (range; 0.05 – 0.80 cm) and 0.30 cm (range; 0.17-0.30 cm), respectively. Neither volume of tissue excised prior to CSM nor that of the CSMs was correlated with occult disease (p=0.52 and p=0.80, respectively). Patient age, race, tumor size, grade, hormone receptor status, LVI, lymph node status, and use of neoadjuvant chemotherapy similarly could not predict occult disease (all p>0.05; see table for selected parameters). However, the presence of invasive lobular cancer (ILC) trended towards association with the finding of occult disease in initially margin-negative patients (p=0.06).

Conclusion: Nearly 12% of patients with negative margins at PM will have occult disease; patients with ILC are more likely to have such occult disease. The impact of such disease on locoregional recurrence, however, remains to be elucidated.

13.14 Intraoperative Blood Loss: Impact on Long-Term Outcomes After Colorectal Liver Metastases Resection

Posted on January 18, 2016

G. MARGONIS1, Y. Kim1, F. Gani1, T. M. Pawlik1 1Johns Hopkins University School Of Medicine,Surgery,Baltimore, MD, USA

Introduction: The influence of intraoperative blood loss (IBL) on long-term outcomes of patients undergoing liver resection for colorectal cancer liver metastases (CRLM) remains not well defined. We sought to study the prognostic impact of intraoperative blood loss on long-term survival following resection of CRLM.

Methods: A total of 433 patients who underwent hepatic resection with curative intent for CRLM between 2000 and 2013 at a major hepatobiliary center were identified. Demographics data, operative details, intraoperative blood loss data, and long-term outcomes were collected and analyzed. IBL cutoff volume was calculated using chi square test analysis. Clinicopathologic predictors of IBL were identified using logistic regression. Overall survival (OS) was assessed using the Kaplan-Meier and Cox regression methods.

Results:Median patient age was 54 (IQR 44, 64) years, and the majority of patients were male (58.9%, n=255). At the time of surgery, the median IBL was 400 (IQR 200-800) mL. Intraoperatively, 146 (33.7%) patients had an IBL <250 mL, while 287 (66.3%) patients had an IBL ≥250 mL. On multivariate analysis, factors associated with IBL ≥250 mL included male sex (OR 2.62, 95%CI 1.69-4.08; P<0.001), tumor size >3cm (OR 1.88, 95%CI 1.18-2.99; P=0.01), and major hepatic resection (OR 3.07, 95%CI 1.93-4.90; P<0.001). At a median follow-up time of 30.6 months, median and overall 5-year survival were 59.9 months and 49.3%, respectively. Of note, IBL was associated with both median and 5-year survival (<250 mL: 70.5 months, 62.0% vs. 251-1000 mL: 56.4 months, 46.1% vs. >1000 mL: 36.9 month, 33.0%, respectively; P=0.004, Figure). On multivariable analysis, tumor specific factors such as primary tumor N stage (HR 1.42, 95%CI 1.04-1.95; P=0.03) and tumor size >3 cm (HR 1.48, 95%CI 1.11-1.98; P=0.01), as well as procedure factors such as use of ablation (HR 2.31, 95%CI 1.59-3.34; P<0.001) were associated with overall survival. Of note, IBL also remained an independent prognostic factor of long-term survival even after controlling for whether the patient did or did not receive a blood transfusion (HR 1.48, 95%CI 1.06-2.07; P=0.02).

Conclusion:The magnitude of IBL during CRLM resection was related to biologic characteristics of the tumor as well as the extent of surgery. Increased IBL during CRLM resection was an independent prognostic factor associated with a worse long-term survival.

13.15 Predicting The Need For Perioperative Transfusion In Liver Surgery

Posted on January 18, 2016

M. Mavros1,2, A. Ejaz3, Y. Kim2, F. Gani2, T. M. Pawlik2 1MedStar Washington Hospital Center,Surgery,Washington, DC, USA 2Johns Hopkins University School Of Medicine,Surgery,Baltimore, MD, USA 3University Of Illinois At Chicago,Surgery,Chicago, IL, USA

Introduction: Blood loss and transfusion have traditionally been a concern when performing a hepatic resection. While many patients will have blood products crossmatched preoperatively, only a proportion will get transfused. We sought to create a score to predict need for transfusion.

Methods: Patients in the 2010-2013 American College of Surgeons National Surgical Quality Improvement Program undergoing liver surgery were analyzed. Multivariable models were constructed to identify independent predictors of perioperative transfusion (≥1 unit PRBCs intraoperatively or within 72 hours postoperatively). A scoring system to estimate odds of transfusion was constructed (n=16,679) and then validated (n=8,169).

Results: Among 24,848 cases analyzed, median age was 60 years and 52% were female. 9001 patients (36%) had a major hepatectomy and 6100 (25%) received a transfusion. Factors predictive of transfusion included preoperative hematocrit (OR 2.4), preoperative transfusion (OR 3.2), major hepatectomy (OR 1.6), extrahepatic surgery (OR 1.3), bleeding disorder (OR 1.8), ASA class (ASA 3-4 OR 1.3, ASA 5 OR 2.1), preoperative albumin, (OR 1.4) and alkaline phosphatase (OR 1.4). A weighted integer score was derived using these factors, which could predict with moderate accuracy the need for transfusion in the validation dataset: score 1 (reference): 9% likelihood of transfusion; score 2: 18%, OR 2.3; score 3: 28%, OR 4.1; score 4: 42%, OR 7.8; score 5: 66%, OR 20.1; AUC: 70.1%.

Conclusion: Up to 1 in 4 patients undergoing hepatic resection required a transfusion. A score derived from preoperative factors including patient comorbidities, laboratory values, and extent of surgery was associated with the need for transfusion.

13.16 Frailty is a Predictor of Postoperative Morbidity and Mortality after Colectomy for Cancer

Posted on January 18, 2016

O. Trofymenko1, R. Venkat1, E. Telemi1, V. Nfonsam1 1University Of Arizona,Surgery,Tucson, AZ, USA

Introduction: Colorectal cancer is the third most common cancer in men and women in the U.S. and is the second leading cause of cancer death in both sexes. Colectomies play an important role as a treatment option for people with colorectal malignancy. Frailty has been noted as a powerful predictive preoperative tool for 30-day postoperative complications.

Methods: The NSQIP participant use file was queried to identify 26,314 patients with malignant colorectal neoplasm. 52.29% underwent laparoscopic (n = 12,555) and 47.71% open (n = 13,759) colectomies. A previously described and validated modified frailty index (mFI) was calculated on the basis of NSQIP variables. Our primary outcomes were overall morbidity, Clavien class IV (requiring critical care support) and Clavien class V (mortality) complications.

Results:Median age was 69 years, and BMI was 27.0Kg/m2. 50.3% of patients were males. 56.0% of patients were ASA Class 3 or higher. The median mFI was 0.09 (0 – 0.73). As mFI increased from 0 (non-frail) to 0.36 and above, the overall morbidity increased from 14.6% to 39.3% (p<0.01) and serious morbidity increased from 6.8% to 22.8% (p<0.01), respectively. The Clavien IV complications rate increased from 2.4% to 16.8% and the mortality rate increased from 0.7% to 7.8%, respectively (p<0.01). On multivariate analysis mFI was independent predictor of serious morbidity (Adjusted Odd Ratio (AOR): 6.2, p<0.01), overall morbidity (AOR: 5.0, p<0.01), Clavien IV complication rates (AOR: 13.0, p<0.01) and mortality (AOR: 4.4, p<0.01).

Conclusion:A simplified frailty index, obtained by easily identifiable patient characteristics, significantly predicts morbidity and mortality after colectomy for cancer. Assessment of frailty may facilitate perioperative risk stratification, as well as help identify and counsel high-risk patients.

13.11 Modern Trends of Paget Disease of the Breast: A Report from the National Cancer Database

Posted on January 18, 2016

A. Hanna1, D. Goto2, S. Kesmodel1, K. Tkaczuk1, S. Feigenberg1, D. Mullins2, N. Hanna1, E. Bellavance1 1University Of Maryland,School Of Medicine,Baltimore, MD, USA 2University Of Maryland,School Of Pharmacy,Baltimore, MD, USA

Introduction: Paget disease (PD) of the breast represents a minority (~2%) of breast malignancies. Patients typically present with eczematous changes of the nipple due to intraepidermal spread of tumor cells. While the traditional treatment of PD is total mastectomy, recent small studies have supported treatment with breast conservation (BC) followed by adjuvant radiation. In this study, we aim to identify the current presenting and management patterns of patients being treated with PD compared to non-PD breast cancer.

Methods: The National Cancer Data Base was used to identify stage 0 – III unilateral breast intraductal or lobular carcinoma female patients 18 – 90 years old who underwent surgery from 2010 to 2012. Patients with PD were compared to patients without PD across demographic disease, and treatment variables. All statistical analyses were two-sided t-tests for continuous variables and chi-squared tests for discrete variables.

Results: Of 502,833 female patients in the database who had surgery in 2010 or later, 2708 PD patients and 284167 non-PD patients met inclusion criteria. PD was associated with underlying ductal carcinoma in situ in 24% and invasive cancer in 76% (78% intraductal and 22% lobular) of cases. Paget disease patients were younger, more likely to be African American, less educated and with lower income, possess government insurance plans (Medicare or Medicaid) and have more comorbidities than non-PD patients (all P < 0.0001). Breast cancer patients presenting with PD were more likely to present at a higher stage, undergo chemotherapy, and start treatment later than non-PD patients (all P < 0.0001). Breast cancers associated with PD were more likely to have lymphovascular invasion and test ER-, PR-, and Her2+. Only 23% of PD patients underwent BC compared with 58% of non-PD patients (p < 0.0001).

Conclusions: There are significant differences in the clinical presentation and local and systemic management of Paget disease of the breast and non-Paget disease in the United States, including disparities based on socioeconomic factors. Despite data supporting the use of breast conservation in the treatment of patients with Paget disease, a minority of these patients are actually treated with breast conserving surgery. Further studies are required to fully understand the extent to which differences in socioeconomic factors and clinical presentation affect the treatment and outcomes of Paget disease patients.

13.12 Determining the Optimal Number of Lymph Nodes Harvested During Esophagectomy

Posted on January 18, 2016

A. Salem1, R. Shridhar2, K. Almhanna2, S. Hoffe2, E. Toloza2, K. Meredith3 3Florida State University College Of Medicine,Department Of Surgery,Tallahassee, FL, USA 1University Of Wisconsin,Department Of Surgery,Madison, WI, USA 2Moffitt Cancer Center And Research Institute,Department Of Surgery,Tampa, FL, USA

Introduction:
Tumor depth and nodal status are critical for adequate staging of esophageal cancer. Assessment of a higher number of lymph nodes (LN) during curative resection has been shown to correlate with improved survival for many cancer types. We sought to examine the impact of the number of lymph nodes (LN) harvested during esophagectomy on esophageal cancer related outcomes.

Methods:
From a comprehensive esophageal cancer database we identified patients who underwent curative resection from 1994 to 2013. The impact of total lymph nodes (LN) retrieved on disease-free survival (DFS) and overall survival (OS) was investigated.

Results:
In total, 635 patients were identified. Patients were divided on the basis of total number of lymph nodes removed (<8, 9-12, 13-20, and >20). There was no difference in the pretreatment distribution of tumor stages among the studied groups (p=0.23). The 5-year OS and DFS rates for the group by lymph node category were (43%, 42%, 55%, and 36%, p=0.18) and (44%, 37%, 46%, and 36%, p=0.52), respectively. Total number of lymph nodes assessed did not correlate with reduced risk of recurrence or improved survival. On multivariable analysis controlling for age, sex, histology, neoadjuvant therapy, only removal of 13-20 LNs correlated to improved oncological overall survival outcome (HR=0.65, p=0.04, CI=0.4-0.9) (Table 1.).

Conclusion:
In a tertiary cancer center, we demonstrated that only removal of 13-20 lymph nodes during esophagectomy correlated to improved overall survival. While the importance of standardized pathologic examination and adequate nodal staging is of utmost importance for patients with esophageal cancer undergoing esophagectomy the optimum number of lymph nodes removed clearly warrants further investigation.

13.13 Neoadjuvant Chemotherapy Changes Breast Surgery Patterns in Early Stage Breast Cancer

Posted on January 18, 2016

O. Kantor1, C. Pesce2, C. H. Wang3, E. Liederbach2, D. J. Winchester2, K. Yao2 1University Of Chicago,Department Of Surgery,Chicago, IL, USA 2Northshore University Health System,Department Of Surgery,Evanston, IL, USA 3Northshore University Health System,Center For Biomedical Research Informatics,Evanston, IL, USA

Introduction:
The rates of neoadjuvant chemotherapy (NAC) are increasing in the US. We hypothesized that more women are undergoing bilateral mastectomy (BM) after NAC then similar women not undergoing NAC for early stage breast cancer.

Methods:
The National Cancer Data Base was used to identify women with invasive, clinical stage I-II breast cancer that underwent either neoadjuvant or adjuvant chemotherapy from 2006-2012.

Results:

262,732 patients with clinical stage I-II breast cancer that had either NAC or adjuvant chemotherapy were identified. Of these, 45,972 (17.5%) had NAC prior to surgery – 5.5% of clinical stage I tumors and 27.6% of clinical stage II tumors. BM rates amongst those with cT1, cT2 and cT3 tumors were 23.4%, 20.8%, and 22.0%, respectively for those undergoing NAC compared to 13.6%, 15.8%, and 20.9% for those undergoing adjuvant therapy (p<0.01). From 2006 to 2012, BM rates among cT1 tumors went from 8.8% to 17.9%, for cT2 tumors from 9.7% to 22.4% and for cT3 tumors from 13.1% to 27.4% (all p<0.01). For those patients who had a partial or complete pathologic response after NAC, BM rates increased an additional 6.5% for cT1 tumors (p<0.01), and <1% for cT2 or cT3 tumors (p>0.05).

From 2010-2012, overall BM rates for triple negative tumors was 18.0% and for Her2 positive tumors 18.1%. NAC increased rates of BM most in patients with triple negative tumors (9.2%) compared to 7.5% for Her2 positive tumors and 6.8% for luminal A tumors (p<0.01). On multivariate analysis adjusting for patient, facility, tumor, and treatment factors, the effect of NAC persisted, with NAC increasing the odds of BM in patients with cT1 tumors (OR 1.63, CI 1.52-1.76) and cT2 tumors (OR 1.29, CI 1.22-1.35), but not in cT3 tumors (OR 1.02, CI 0.89-1.18).

Conclusion:
The use of neoadjuvant chemotherapy significantly increases the rate of bilateral mastectomy in early stage breast cancer. This is most pronounced in cT1 tumors amenable to breast conservation.

13.08 ICU Utilization Following Major Cancer Resections Differs Between High and Low Mortality Hospitals

Posted on January 18, 2016

S. L. Revels1, P. K. Park1, J. D. Birkmeyer2, S. L. Wong1 1University Of Michigan,Ann Arbor, MI, USA 2Dartmouth Medical School,Lebanon, NH, USA

Introduction: Case-fatality and overall mortality rates after major cancer resections vary widely across hospitals. The mechanisms that drive these differences in outcomes are poorly defined. We examined the extent to which the utilization of critical care resources explains disparate outcomes between low mortality hospitals (LMHs) and high mortality hospitals (HMHs).

Methods: All hospitals participating in the National Cancer Database (2006-2007) were ranked on risk-adjusted mortality for major bladder, colon, esophagus, gastric, lung and pancreas cancer resections. Onsite chart reviews were performed at 19 LMHs (risk-adjusted mortality rate 2.4%) and 30 HMHs (risk-adjusted mortality rate 6.7%), abstracting information on perioperative care, complications and mortality. Using logistic regression, differences in complication, case-fatality and mortality rates were examined based on ICU utilization. Models were adjusted for patient factors, tumor characteristics and clustering within hospitals.

Results: Overall, 40% of patients were triaged directly to an ICU following major cancer resection. Relatively few patients were transferred to ICUs after POD 0, with rates of 4.4% in LMHs and 3.8% in HMHs (p=0.27). HMHs admitted 45.7% of patients directly to an ICU postoperatively, significantly more than 32.6% at LMHs (p<0.01). HMHs directly admitted 76.6% of patients with an ASA status of 4 or 5, 60.3% of patients with ischemic heart disease and over 80% of patients undergoing lung, esophagus and pancreas resections to an ICU. After risk-adjustment, the complication rate for HMH ICU direct admits was 31.4% (95%CI, 24.4-38.6%), not significantly different from 25.6% (95%CI, 20.1-30.5%) in LMH direct ICU admits. Conversely, the risk-adjusted case-fatality rate was significantly higher for HMH direct ICU admits compared to those in LMHs, 32.8% (95%CI, 24.5-41.2%) versus 13.8% (95%CI, 8.0-19.5%), respectively. Similarly, risk-adjusted overall mortality for ICU direct admits was significantly higher in HMHs than LMHs, 12.6% (95%CI, 10.1-15.1%) compared to 4.8% (95%CI, 3.2-6.4%), respectively.

Conclusion: Preemptive ICU admission does not assure better outcomes for high risk cancer patients. Despite utilizing critical care resources at substantially higher rates for high risk patients compared to LMHs, HMHs experienced significantly higher case-fatality and overall mortality rates. Understanding the processes and structural aspects of ICU care administered to cancer patients may yield opportunities to improve the quality of cancer surgery.

13.09 Trends in Neoadjuvant Therapy Utilization and Short-term Outcomes in Resected Pancreatic Cancer

Posted on January 18, 2016

K. A. Mirkin1, C. Hollenbeak1, N. Gusani1, J. Wong1 1Penn State Hershey Medical Center,General Surgery,Hershey, PA, USA

Introduction:
Pancreatic cancer is the fourth leading cause of cancer death in the US with surgical resection and chemotherapy offering the only chance of long-term survival. Neoadjuvant therapy (NAT) is utilized in certain patients to optimize outcomes. The objective of this study was to analyze trends in NAT utilization and compare short-term outcomes in patients receiving NAT and surgery to patients undergoing initial surgical management for clinical stage I – III pancreatic cancer.

Methods:
This was a retrospective analysis of the National Cancer Data Base (2003-2011), evaluating trends in utilization and short-term outcomes (30- and 90-day mortality and hospital length of stay) in pancreatic cancer patients. Patients with clinical stages I-III adenocarcinoma and carcinoma of the pancreas who underwent surgery, both with and without NAT (chemotherapy, radiation, or both), were included. Univariate statistics were used to compare characteristics of patients who underwent initial surgery versus NAT followed by surgery. Logistic and linear regression models were used to model 30- and 90-day mortality and length of stay, respectively.

Results:
The analysis included 6,204 patients who underwent initial surgery and 10,245 patients who received NAT and surgery. Patients undergoing initial surgery tended to be older, female, covered by Medicare, and had lower staged disease. Over the past decade, initial surgical management and multimodality NAT has steadily decreased. Use of neoadjuvant radiation has remained relatively static with low rates of utility. Use of neoadjuvant chemotherapy, however, has been rising steadily. Mortality rates at 30 and 90 days were significantly higher for patients undergoing initial surgery compared to patients receiving NAT and surgery (9.3% vs. 0.5%, p<0.0001, 17.8% vs. 2.0%, p<0.0001, respectively). After controlling for patient, disease and surgery characteristics, patients who received NAT had a 94% lower odds of mortality at 30 days (p < 0.0001) and an 90% lower odds at 90 days (P<0.0001). Length of stay was significantly higher for patients undergoing initial surgery as compared to patients receiving NAT and surgery (13.4 days vs. 10.1 days, p<0.0001). More advanced clinical stage disease, pancreatoduodenectomy or total pancreatectomy also negatively impacted 30 and 90 day mortality as well as hospital length of stay.

Conclusion:
The utilization of neoadjuvant chemotherapy has been increasing steadily for the past decade, while initial surgical resection has been decreasing. Use of multimodality NAT also appears to be decreasing. Neoadjuvant therapy does not appear to have any adverse effect on short-term outcomes, including 30- and 90-day mortality and hospital length of stay.

13.10 Incomplete Neoadjuvant Radiotherapy Is Associated With Inferior Survival For Rectal Cancer Patients

Posted on January 18, 2016

Z. Sun1, M. A. Adam1, J. Kim1, M. Palta2, B. G. Czito2, J. Migaly1, C. R. Mantyh1 1Duke University Medical Center,Surgery,Durham, NC, USA 2Duke University Medical Center,Radiation Oncology,Durham, NC, USA

Introduction:
While failing to complete chemotherapy has been shown to adversely affect survival in colorectal patients, the effect of incomplete neoadjuvant radiotherapy dosing is unclear.

Methods:
Adults with stage II-III rectal adenocarcinoma from the 2006-2012 National Cancer Data Base who received neoadjuvant chemoradiation followed by surgical resections were included. Multivariable regression methods were used to compare resection margin positivity, permanent colostomy rate, 30-day readmission, 90-day mortality, and overall survival between patients who received complete (45-50.4 Gy) and incomplete (<45 Gy) doses of radiation preoperatively.

Results:
Among 18,060 patients included, 971 (5%) received incomplete doses of neoadjuvant radiation. Median radiation dose received among those who did not achieve complete dosing was 36 Gy (Interquartile range 19.8-41.4 Gy). Female sex (OR 0.70, p<0.001) and receiving radiation at a different hospital (OR 0.71, p<0.001) were independent predictors of failing to achieve complete dosing, while private insurance status was predictive of success (OR 1.54, p= 0.003). At 5-years follow up, overall survival was improved among patients who received complete radiotherapy (73% vs. 63%, p<0.001). After adjustment for demographic, clinical, and tumor characteristics, patients receiving complete versus incomplete radiotherapy dose had similar resection margin positivity, permanent colostomy rate, 30-day readmission, and 90-day mortality (all p>0.05). However, complete radiation dose had significantly lower risk of long-term mortality (HR 0.70, p<0.001).

Conclusion:
Achieving target radiation dose of 45-50.4 Gy is associated with survival benefit in patients with locally advanced rectal cancer. Regionalization of multimodality oncology care may increase probability of completing neoadjuvant therapy.

13.05 External Radiation Improves Survival in Margin-Negative Stage II Pancreatic Adenocarcinoma

Posted on January 18, 2016

O. Kantor1, W. Lutfi2, M. S. Talamonti2, C. H. Wang3, D. J. Winchester2, R. A. Prinz2, M. S. Baker2 1University Of Chicago,Department Of Surgery,Chicago, IL, USA 2Northshore University Health System,Department Of Surgery,Evanston, IL, USA 3Northshore University Health System,Center For Biomedical Research Informatics,Evanston, IL, USA

Introduction:
The benefit of adjuvant external beam radiation following a margin negative resection in early stage pancreatic cancer has not been definitively determined.

Methods:
We queried the National Cancer Data Base for patients with pathologic stage I-II pancreatic adenocarcinoma who underwent resection between 1998 and 2011. Patients receiving neoadjuvant therapy and those with R1or R2 resection margins were excluded. Multivariate Cox-regression modeling was used to analyze stage-specific survival.

Results:

25,012 patients with stage I-II pancreatic adenocarcinoma underwent upfront surgical therapy and had a margin negative resection during the period studied. 1683 (6.7%) were pathologic stage IA, 3055 (12.2%) stage IB, 5850 (23.4%) stage IIA, and 14424 (57.7%) stage IIB. 4909 (19.6%) patients received adjuvant chemotherapy, 9367 (37.5%) adjuvant chemotherapy and radiation (chemoRT), and 10736 (42.9%) received no adjuvant therapy. ChemoRT utilization increased with increasing stage (25.3% in stage IA vs 41.0% in stage IIB, p<0.001). ChemoRT was more common at community than academic centers (46.3% vs 33.5%, p<0.001) and more common at low volume than high volume centers (44.6% vs 31.9%, p<0.001).

Cox-regression adjusted for age, race, comorbidities, facility type, location, and volume, type of pancreatectomy, and grade was used to estimate stage-specific survival for patients undergoing treatment between 1998-2006. Treatment at a high volume center was associated with decreased mortality (HR 0.74-0.81, p<0.04) across all stages. Age ≥70 (HR 1.2-1.3, p<0.01) and higher grade (HR 1.55-1.82, p<0.01) were associated with higher risk of mortality at all stages. ChemoRT was associated with a benefit in median overall survival over chemotherapy alone in all stages. That benefit was statistically significant in patients with stage IIA (23.4 vs 19.8 months, p=0.01) and stage IIB (18.6 vs 16.0 months, p<0.01), but not stage IA (37.4 vs 31.2 months, p=0.28) or stage IB (25.9 vs 20.5 months, p=0.06) [Figure 1].

Conclusion:
Addition of radiation to adjuvant chemotherapy after margin negative resection of pancreatic adenocarcinoma provides a survival benefit in patients with pathologic stage II disease and should be considered as adjuvant therapy in these patient groups.

13.06 Pediatric Papillary Thyroid Cancer Patients Referred to High Volume Centers Have Improved Outcomes

Posted on January 18, 2016

L. M. Youngwirth1, M. A. Adam1, S. M. Thomas1, R. P. Scheri1, S. A. Roman1, J. A. Sosa1 1Duke University Medical Center,Durham, NC, USA

Introduction: Papillary thyroid cancer is the most common endocrine malignancy in children, but it is still rare. Therefore, data are limited on patient outcomes. This study sought to determine pediatric papillary thyroid cancer patient outcomes by facility case volume on a national level.

Methods: The National Cancer Data Base (1998-2011) was queried for all pediatric patients (age ≤18) with papillary thyroid cancer. Demographic, clinical, and pathologic features at the time of diagnosis were evaluated for all patients. Facility case volume was defined as the total number of pediatric thyroid cancer patients presenting at a given hospital, and case volume quartiles were created that contained an approximately equal number of facilities. Patients then were placed into quartiles based on these case volume definitions. Univariate analysis was performed to measure association with outcomes, including length of stay and readmissions.

Results: In total, 3,809 patients met inclusion criteria. The majority were female (80.2%) and white (89.2%). The mean tumor size was 2.6 ± 2.9 cm, and 49.7% of patients had metastatic disease in cervical lymph nodes. The majority of patients underwent total thyroidectomy (88.7%) and received radioactive iodine (59.7%). The five year overall survival was 99.0%. Patients treated at the highest volume facilities were younger (mean age of 14.8 years vs 16.0 years), more likely to be treated at an academic facility (71.7% vs 3.0%), and more likely to travel a greater distance to their treatment facility (mean distance of 50.1 vs 13.8 miles) (all p <0.01). Also, patients treated at the highest volume facilities were more likely to undergo total thyroidectomy (91.0% vs 81.5%) and receive radioactive iodine (63.1% vs 54.2%) (all p <0.01). The mean length of stay was similar in all quartiles (p =0.71); however, patients treated at the highest volume facilities were less likely to be readmitted than patients treated at the lowest volume facilities (5.2% vs 9.0%; p <0.01).

Conclusion: Pediatric patients with papillary thyroid cancer treated at low volume centers were more likely to be readmitted after thyroid surgery than patients treated at high volume centers. Providers should consider facility case volume status when referring these children for thyroid surgery.

13.07 Minimally Invasive Adrenalectomy for Adrenocortical Cancer: An Analysis of 506 Patients

Posted on January 18, 2016

M. A. Adam1, S. A. Roman1, J. A. Sosa1,2 1Duke University Medical Center,Department Of Surgery,Durham, NC, USA 2Duke Clinical Research Institute,Durham, NC, USA

Introduction:
Controversy exists regarding the role of minimally invasive adrenalectomy (MIA) for adrenocortical carcinoma (ACC) due to concerns regarding adequacy of oncologic resection. Published data are limited by small sample size. This study examines patterns of use, completeness of resection, and survival associated with utilization of MIA vs. open adrenalectomy for ACC.

Methods:
Adult patients with ACC undergoing surgery were identified from the National Cancer Data Base (2010-2012). Multivariable regression and survival models were used to examine impact of surgical approach on margin status and survival while adjusting for tumor size, extra-adrenal extension, and extent of surgery.

Results:
506 ACC patients were included; 151 (30%) MIA, and 355 (70%) open surgery. Non-academic centers performed 47% of MIA and 38% of open cases. Open adrenalectomy was performed more often if there was a preoperative suspicion/diagnosis of ACC (82% vs. 18% for MIA, p<0.01). The proportion of unsuspected ACC cases undergoing MIA was 100% at community centers, 74% at community comprehensive centers, and 72% at academic centers (p=0.2). There were 18 cases converted from MIA to open surgery; 14 (78%) were unsuspected ACC. Compared to open adrenalectomy, patients undergoing MIA had smaller (13 vs. 7 cm) and more intra-adrenal tumors (46% vs. 60%), (all p<0.01). Hospital length of stay was shorter for MIA vs. open surgery (4 vs. 6 days, p<0.01), while overall rates of positive margins were similar (18% vs. 18%). After adjustment, length of stay remained shorter for MIA (-2 day, p<0.01); however, surgery at non-academic centers was associated with increased odds of positive margins (OR 1.94, p=0.03) compared to academic centers. Median follow-up was 22 months. After adjustment, overall survival was similar between MIA and open adrenalectomy (HR 1.10, p=0.65).

Conclusion:
MIA is being used for ACC in approximately a third of patients, with the majority of cases not suspected to be malignant preoperatively. While MIA is associated with a similar risk of positive margins and survival as open adrenalectomy, non-academic centers have a nearly double rate of incomplete resection. Our data emphasize the importance of preoperative evaluation and adequate referral for patients with large adrenal masses.

12.20 Learning From Failure in The Modern Era: Results of Reoperative Surgery After Failed Parathyroidectomy.

Posted on January 18, 2016

S. Zaheer1, L. Kuo1, H. Wachtel1, R. Roses1, G. Karakousis1, R. Kelz1, D. Fraker1 1Hospital Of The University Of Pennsylvania,Department Of Surgery,Philadelphia, PA, USA

Introduction:

Parathyroidectomy is a common surgical procedure. Currently, 10% of surgeons practice bilateral neck exploration(BNE), 68% use a minimally invasive approach(MIP), and 22% have a mixed practice. We sought to determine the reasons for failure of MIP as compared to failure of BNE.

Methods:

Patients with primary hyperparathyroidism who underwent reoperative surgery for recurrent or persistent disease were identified in our institutional prospective endocrine surgery registry(1997 to 2013) for the study. The primary outcome of interest was reason for failure of the initial surgery determined by intra-operative findings and pathologic review. Univariate analysis was performed to examine differences across reoperative patients who underwent BNE compared to those who underwent MIP as the initial procedure using the Chi-square test, as appropriate.

Results:

Of 141 patients who met study criteria, 28% (n=39) had undergone MIP and 72% (n=102) BNE. Patient characteristics including age, race, gender, body mass index, calcium, PTH level and symptoms were similar across the two groups. Recurrent disease was associated with abnormal glands in ectopic locations (n=45), adenomas in normal positions(n=33), hyperplasia(n=25), MEN-associated hyperplasia(n=21), parathyroid cancer(n=6), abnormal supernumerary glands(n=5), and parathymatosis (n=4). Reoperation failed to identify the cause of failure in 2 patients. A single parathyroid adenoma was the most common cause of operative failure following MIP. Adenoma in a normal location was more common following MIP than BNE (43.59% vs 15.09%, p=0.005). In BNE failures, ectopic gland was the most common cause. The frequency of failure due to ectopic gland was greater amongst BNE when compared to MIP(31 % vs. 7%, p=0.005). In initial MIP, IOPTH monitoring was used in 45% (n=17) of patients. Of these patients, 4 had intraoperative parathyroid hormone (IOPTH) levels which failed to normalize; 3 had incorrect interpretation of IOPTH.

Conclusion:

Failure following MIP is often due to a missed adenoma in a normal location as opposed to failure following BNE, which is often associated with a missed ectopic gland. All patients should be counseled on the possibility of an ectopic gland prior to parathyroid exploration. Experienced surgeons preferably skilled in the use of IOPTH monitoring should perform MIP to avoid unnecessary failure.

13.02 Insurance Status, Not Race, is Associated with Use of Minimally Invasive Approach for Rectal Cancer

Posted on January 18, 2016

M. C. Turner1, M. A. Adam1, Z. Sun1, J. Kim1, B. Ezekian1, C. Mantyh1, J. Migaly1 1Duke University Medical Center,Surgery,Durham, NC, USA

Introduction:
Minimally invasive surgery (MIS) is increasingly utilized in rectal resection for adenocarcinoma given the superior short-term and equivalent oncologic outcomes. Race and socioeconomic status have been implicated in disparities of rectal cancer treatment and resection. Our aim was to determine the impact of patient race and medical insurance on use of minimally invasive (MIS) compared to open techniques for rectal cancer resections in the United States.

Methods:
Adult patients undergoing MIS rectal resections (laparoscopic or robotic) for stage I-III rectal adenocarcinoma were included from the National Cancer Data Base (2010-2012). Multivariate analyses were employed to identify factors independently associated with use of MIS vs. open resection.

Results:

A total of 23,274 patients were included: 9,164 (39%) underwent MIS and 14,110 (61%) open surgery. Overall, 86% were white, 8% black, 3% Asian, and 2% identified as ‘other.’ In unadjusted analysis, factors associated with use of open vs. MIS were black race, Medicare/Medicaid insurance, and lack of medical insurance. However, after adjustment for patient demographic, clinical, and treatment characteristics, black race was not associated with use of MIS vs. open surgery (OR 0.90, p=0.07). With adjustment, compared to privately insured patients, uninsured patients (OR 0.52, p<0.01) and those with Medicare/Medicaid (OR 0.79, p<0.01) were less likely to receive minimally invasive resections. Lack of insurance was significantly associated with less frequent use of MIS (OR 0.59, p=0.02) in black patients and (OR 0.51, p<0.01) in white patients. Across all uninsured patients, black race was not associated with lower frequency of MIS approach (OR 0.96, p=0.59).

Conclusion:

Medical insurance status, not race, is associated with utilization of minimally invasive techniques for oncologic rectal resections. In light of the short term benefits and cost-effectiveness, strategies to improve access to minimally invasive techniques in this patient population should be oriented toward expanding insurance coverage.

13.03 A National Study of Outcomes Associated with Conversion in Laparoscopic Colectomy for Colon Cancer

Posted on January 18, 2016

B. Yerokun1, M. Adam1, Z. Sun1, J. Kim1, S. Sprinkle1, J. Migaly1, C. Mantyh1 1Duke University Medical Center,Department Of Surgery,Durham, NC, USA

Introduction: Limited data exist with regard to the effect of conversion from laparoscopic to open colectomy on oncologic outcomes in colon cancer. The aims of this study were to evaluate perioperative and oncologic outcomes associated with conversion of laparoscopic to open colectomy in the United States.

Methods: The National Cancer Data Base was used to identify patients who underwent colon resection for Stage I-III colon cancer (2010-2012). Patients were stratified into three groups: laparoscopic colectomy (LC), converted laparoscopic to open colectomy (CC), and open colectomy (OC). Multivariable modeling was applied to compare perioperative and oncologic outcomes from CC and LC to open colectomy while adjusting for patient, clinical, and tumor characteristics.

Results: A total of 104,400 patients were included: 40,328 (38.63%) had LC, 6,144 (5.89%) CC, 57,928 (55.48%) OC. The conversion rate was 13.2%. After adjustment, surgical margin status was not significantly different between CC and OC (OR 1.05, 95% CI 0.93-1.19, p=0.44). However, with adjustment, CC was associated with shorter hospital length of stay (-4%, 95% CI -2% to -5%, p<0.0001) and lower odds of 30-day mortality (OR 0.77, 95% CI 0.64-0.94, p=0.01) when compared with OC. In adjusted survival analysis with median follow-up time of 24 months (range 1-51 months), overall survival was similar between the CC and open groups (HR 0.96, 95% CI 0.88-1.05, p=0.34). Compared with OC, LC was associated with improved 30-day mortality (OR 0.47, 95% CI 0.42-0.52, p<0.0001) and overall survival (HR 0.69, 95% CI 0.66-0.73, p<0.0001) after adjustment.

Conclusion: In this nationally representative study, patients who underwent initial laparoscopic attempts at colon resection for non-metastatic colon cancer before conversion to open colectomy had similar oncologic outcomes, and improved short-term outcomes, when compared with standard open colectomy. This analysis demonstrates that laparoscopic colectomy should be considered on all patients without contraindication to laparoscopy, and conversion can be done safely when necessary, without harm to patients.

13.04 Gastrointestinal Neuroendocrine Carcinoma is an Increasingly Common Cancer and Cause of Death

Posted on January 18, 2016

A. K. Idicula2, S. Dissanaike2, M. S. Wachtel1 2Texas Tech University Health Sciences Center,Surgery,Lubbock, TEXAS, USA 1Texas Tech University Health Sciences Center,Pathology,Lubbock, TX, USA

Introduction:
Neuroendocrine carcinoma is not uncommonly seen in the stomach and more distal gastrointestinal tract, but its relative incidence and mortality among the several primary sites has not been fully evaluated. Neither have secular trends been compared. We hypothesized that examining a large database would provide better understanding of these tumors as respects the different gastrointestinal sites.

Methods:
Patients were included if they had histologically proven neuroendocrine carcinoma (ICD-O-3 8002, 8013, 8041-8045, 8240-8246, 8249), if they were at least 20 years old, and if the primary site was the stomach (ICD-O-3 C16.1-C16.9), the small bowel (ICD-O-3 C17.1-C17.9), the appendix (ICD-O-3 C18.1), or the large intestine (ICD-O-3 C18.0, C18.2-C18.9, C19.9, C20.9). SEER*Stat 8.2.1 queried SEER 13 registries for 1988-2012 incidence and incidence based mortality rates and standard errors. Joinpoint 4.1.2 calculated estimates and standard errors of annual percent changes (APC) and joinpoints and performed tests of parallelism.

Results:

Evaluated were 17,197 cases of and 7,054 deaths. The figure shows incidence and mortality rates and APC by primary site. For all sites, mortality and incidence increased between 1988 and 2012. The least common source of tumors and tumor deaths was the stomach. The most common source of tumors was the large intestine. The most common source of tumor deaths was the small intestine. Parallelism tests showed for incidence, only the stomach and the large intestine (P = 0.53) bore similar APC; for mortality, only the stomach and the appendix (P = 0.83) and the stomach and the large intestine (P = 0.72) bore similar APC. Apart from a non-statistically significant decline in large intestinal incidence rates, APC -4.4 (95% CI -15.0, 7.5), rates were increasing at periods ending in 2012: for incidence, stomach APC 3.3 (95% CI 1.2, 5.4), small intestine APC 3.7 (95% CI 3.3, 4.1), appendix APC 7.2 (95% CI 5.7, 8.7); for incidence based mortality, stomach APC 4.2 (95% CI 2.8, 5.6), small intestine APC 2.1 (95% CI 1.3, 2.8), appendix APC 4.4 (95% CI 2.5, 6.2) , large intestine APC 11.3 (95% CI 1.8, 21.7).

Conclusion:

Neuroendocrine carcinoma of the stomach, the small intestine, the appendix, and the large intestine shows increasing incidence and mortality rates. Rates and rate changes in general differed among sites. Results support the notion of gastrointestinal diversity as respects neuroendocrine carcinoma.

12.17 Understanding the Clinical Implications of Resident Involvement in Uncommon Operations

Posted on January 18, 2016

S. Dasani1, K. D. Simmons1, E. Bailey1, R. Hoffman1, K. Collier1, R. R. Kelz1 1University Of Pennsylvania,Department Of Surgery,Philadelphia, PA, USA

Introduction: Uncommon operations are defined by the Surgical Council on Resident Education as rare, often urgent, operations. Outcomes of these procedures remain largely understudied. Uncommon operations may confer greater operative risk than complex operations and have the educational challenge of occurring infrequently. This study examines the incidence of post-operative events and the role of the resident following uncommon operations.

Methods: We identified uncommon general surgical operations using the ACS National Surgical Quality Improvement Program (NSQIP) Participant Use file (2008-2011). Death or serious morbidity (DSM) within 30 days of the operation was the primary outcome of interest. Failure to rescue (FTR) and prolonged operative time (PRopt) were examined as secondary outcome measures. PRopt was defined by procedure type as ≥ to the 90%ile of operative time. Resident participation was defined as resident involved (RI) or no resident involved (NRI), and stratified by post-graduate year (PGY): 1-3, 4-5, and 6+. Independent multivariate logistic regression models were developed to examine the association between any RI or PGY and outcomes.

Results: Resident participant data was available for 21,453 (84.5%) of the uncommon operations with NRI in 25.4% (5,447) and participation by a PGY1-3 in 12.6% (2,699), a PGY4-5 in 50.4% (10,817), and a PGY6+ in 11.6% (2,490). The observed DSM rate was 28.6% and the observed FTR rate was 5.8%. Results of binary analyses are displayed in the prose and PGY results are seen in the table. Overall, there was no difference in DSM by RI status (NRI: 1,528; 28.1% vs RI: 4,602; 28.8%; p=0.324); however, the PGY level was associated with DSM (Table 1). Any RI was associated with a lower rate of FTR (5.1%) when compared to NRI (8.3%, p<0.001) with decreasing FTR events by increasing PGY (Table 1). After adjustment for patient risk factors, any RI remained associated with a lower likelihood of FTR than NRI (OR: 0.66, 95% CI: 0.50-0.89) with decreasing likelihood of FTR by increasing PGY (Table 1). RI was associated with PRopt in univariate and multivariate analyses (Table 1).

Conclusion: Uncommon operations were associated with substantial DSM. The involvement of PGY4-5 residents was associated with the greatest likelihood of DSM. Regardless of the PGY of the involved resident, all cases with RI demonstrated a lower likelihood of risk-adjusted FTR. The explanation for these findings is not clear; however, the involvement of more senior residents in the technical aspects of uncommon operations may lead to improved results.

12.18 Combining Loss of Muscle Mass and Muscle Attenuation to Predict Outcomes following HPB Surgery

Posted on January 18, 2016

L. Xu1, Y. Kim1, F. Gani1, G. A. Margonis1, D. Wagner1, S. Buttner1, T. M. Pawlik1 1Johns Hopkins University School Of Medicine,Surgery,Baltimore, MD, USA

Introduction: Skeletal muscle depletion (SMD) has been shown to be a powerful predictor of a poor prognosis. We sought to identify the prevalence of sarcopenia and low muscle attenuation (MA) among patients undergoing hepato-pancreatico-biliary (HPB) surgery, as well as the prognostic value of SMD for HPB surgery.

Methods: Patients undergoing HPB surgery between August 2011 and June 2014 with available preoperative (≤30 days) lumbar computed tomography (CT) images were identified. Total psoas volume (TPV) and average psoas density (PD) were measured using preoperative CT scans. Sarcopenia was defined as the lowest gender-specific quartile for TPV. Similarly, low MA was defined as the lowest gender-specific quartile for PD. SMD was defined as presence of both sarcopenia and low MA. Patients with missing data for TPV, PD, or BMI, and patients <18 years were excluded from the analysis. Clinical features, complications, short-term outcomes and overall survival of patients were collected.

Results: Of the 913 patients included, the median age was 63 years (IQR 53, 71) with 47.3% being male. Over two-thirds (n=633, 69.3%) of patients underwent surgery for a malignant disease. Patients undergoing surgery for a malignant disease were older (median 64 years vs. 59 years, p<0.001), and more likely to be male (57.2% vs. 42.5%, p<0.001). Of note, BMI was not different between patients undergoing surgery for malignant or benign disease (p=0.682). Sarcopenia and low MA were more common in patients with malignant disease (sarcopenia 27.7% vs. 19.3% in benign, p=0.007; low MA 27.5% vs. 19.3%, p=0.008). Among the entire cohort, patients presenting with SMD reported a higher incidence of postoperative complications (31.6% vs. 15.8% in non-SMD patients, p<0.001), as well as longer length of stay (median 11 days [IQR 7, 16] vs. 8 days [IQR 6, 12] in non-SMD patients, p<0.001). Among the patients undergoing surgery for malignant disease, patients with SMD had a higher risk of death than patients without SMD (HR 1.8, 95% CI 1.1-2.9; p=0.01). On multivariate analysis, SMD was remained as an independent predictor for both complication (OR 1.90, 95% CI 1.14-3.15; p=0.01) and a worse overall survival (HR 1.66, 95% CI 1.04-2.67; p=0.03).

Conclusions: Sarcopenia and low MA are more common in patients undergoing HPB surgery for malignant disease compared to patients with benign disease. SMD is an independent predictor of poor prognosis in patients undergoing HPB surgery.

12.19 Accidental Punctures and Lacerations in Hepatobiliary Operations: Two Methods of Analysis

Posted on January 18, 2016

A. Shmelev1, A. M. Sill1, K. W. Shaw1, G. C. Kowdley1, J. A. Sanchez1, S. C. Cunningham1 1Saint Agnes Hospital Center And Cancer Institute,Surgery,Baltimore, MD, USA

Introduction: Accidental punctures and lacerations (APL) during the surgical care of hepatobiliary (HPB) patients are not infrequent and are often preventable. The Agency for Healthcare Research and Quality (AHRQ) Patient Safety Indicator #15 (PSI-15) was created to determine rates of APL. Unfortunately, using the indicator alone may be insufficient to fully understand the nature of these events in a large dataset. Another approach to APL rates is using diagnostic and procedural codes (ICD-9), which may identify these events indirectly. We hypothesized that these two approaches, PSI-15 and a composite measure of relevant ICD-9 codes, would yield different incidences of APL, and that this information can provide a more accurate picture of the frequency and characteristics of APL events.

Methods: Using the National Inpatient Sample from the Healthcare Cost and Utilization Project for 2000–2012, we analyzed all cholecystectomies plus pancreatic and liver resections. All PSI-15 codes were examined and rates were calculated using AHRQ software. Cases were divided in to two groups, those carrying the PSI-15 marker of APL, and those not. We then determined the frequency of all procedural and diagnostic ICD-9 codes shared by both groups, and selected those codes with the most discordance. Pairwise comparisons (chi-square tests) of each selected code against PSI-15 as a surrogate of APL presence were determined. Diagnostic (n=54) and procedural (n=12) codes that significantly increased the odds of having a PSI-positive finding (with 95% CIs that did not include 1) were selected for inclusion into one composite measure (CM) of APL. Both CM and PSI-15 were plotted along a timeline. Seasonal trends decomposition and log-linear Poisson regression analyses were carried out to test for observed significance of trends over time.

Results:The rate of PSI-15 varied from 1.2% in 2000 to a flat maximum of 1.5% in 2007 and back to 1.1% in 2012. Rates of the composite marker gradually increased from 3.1% in 2000) to plateau at 3.7% in 2011. A Poisson regression analysis of both trends, for PSI-15 and for CM, demonstrated statistical significance (p<0.001).

Conclusion:The ICD-9 codes in our CM occurred more frequently than PSI-15. In general, both measures trend similarly over time with CM exhibiting larger variation. The divergence seen between these two curves in 2007 may be associated with known changes in Diagnosis Related Groups at that juncture and, possibly, reimbursement policy involving adverse events. While arguably less specific, CM may increase sensitivity for detecting APL events during HPB operations in comparison to PSI-15. These results may also inform the interpretation of APL studies using either PSI-15 or ICD-9 approaches.

« Previous Page
Next Page »