102.11 Effect of Temporary Shunts on Neurologic Outcomes in Repair of Penetrating Carotid Artery Injuries

Posted on January 31, 2019

S. R. Brown1, A. D. Person1, G. E. Mendoza1, M. Dale1, D. Rigg1, D. Keleny1, J. Dabestani1, D. K. Agrawal1, J. A. Asensio1  1Creighton University Medical Center,Surgery,Omaha, NE, USA

Introduction:  The management of penetrating carotid artery injuries continues to evolve. Early questions regarding repair versus ligation have largely been settled in favor of repair for all but the most devastating injuries. To date, the use of temporary shunts to maintain cerebral perfusion in the repair of carotid artery injuries has not been well studied. Thus far, no single study has had the power to state whether the use of shunts makes any difference in neurological outcomes in the setting of penetrating carotid artery injuries.

Methods: A focused literature search was performed with PubMed, Scopus, and Ovid. PRISMA guidelines were followed. Inclusion criteria were those series reporting surgical repair for penetrating carotid artery injuries. Exclusion criteria were endovascular repair, neurosurgical techniques involving craniotomy, and studies with no reports on preoperative or postoperative neurologic status. Further relevant studies found as references in these papers were also evaluated. A total of 28 papers were identified that met these criteria, ranging from 1970 to 2012. Studies were systematically analyzed to extract cases of surgical repair for which both the utilization of temporary shunts and perioperative neurologic outcome data was reported. Non-parametric data was analyzed with Fisher Exact or Chi-Square tests as applicable. Statistical significance was set to a p-value less than 0.05.

Results: There were a total of 717 carotid artery injuries. 592 were repaired without shunts. 125 were repaired with shunts. There was a trend towards lower overall mortality in repairs with shunts vs. those without shunts (5.6% vs. 9.8%; p=0.17). There was a trend towards increased rates of neurological improvement in repairs with shunts vs. those without shunts (13.6% vs. 10%, p=0.34). Patients undergoing repairs with shunts were more likely to have improved or unchanged neurologic outcomes in comparison to those without shunts (92% vs. 84%, p=0.018).

Conclusion: This focused analysis offers the first large scale evidence that the use of temporary shunts in the surgical repair of penetrating carotid artery injuries results in better neurologic outcomes in comparison to repair without the use of shunts. Based on this evidence, we recommend the routine use of temporary shunts during complex repair of penetrating carotid artery injuries.

 

102.10 Fibrinolysis Spectrum in the Burn Population

Posted on January 31, 2019

H. B. Cunningham1, L. R. Taveras1, M. L. Pickett1, J. B. Imran1, T. D. Madni1, S. Park1, M. Zhou1, F. M. Adeyemi1, H. A. Phelan1, M. W. Cripps1  1University Of Texas Southwestern Medical Center,Surgery,Dallas, TX, USA

Introduction: Viscoelastic testing is regarded as superior to conventional clotting assays in detecting coagulation dysfunction after burn injury. A hypercoagulable state has been observed by rotational thromboelastography (ROTEM) in these patients. Moreover, close to a third of patients with severe burns develop disseminated intravascular coagulation. A mortality difference has been found within the fibrinolysis spectrum for trauma patients. The distribution, and associated mortality, of burn patients within this range is unknown. Our aim is to describe the distribution of fibrinolysis, as measured by ROTEM, and identify associated mortality rates.

Methods: All the patients that underwent ROTEM assays were screened at an urban, Level 1 burn center from July 2014 to December 2017. Clinical and ROTEM data were analyzed on burn patients. Data from the initial ROTEM at time of admission was included for evaluation. Hyperfibrinolysis (HF) was defined as maximum lysis on EXTEM >15%. Hypofibrinolysis was defined as maximum lysis on EXTEM <3%.

Descriptive statistics were compared using Fischer’s exact test and the Kruskal-Wallis test for categorical and continuous variables, respectively.

Results: ROTEM results from 1162 patients were reviewed and 116 corresponded to burn patients. Five patients were excluded due to incomplete ROTEM data. A total of 111 patients were included in our study. Median age was 45 years (IQR 33.5 – 58), 79% were male and median TBSA% was 16 (IQR 32.5 – 44). No differences were found in gender, age, ethnicity, race, admission weight, burn type, presence of inhalation injury or presence of concomitant trauma between the groups.

Distribution of fibrinolysis was: hypofibrinolysis, 26 (23.4%); physiologic, 83 (74.8%); and hyperfibrinolysis, 2 (1.8%). Mortality during admission was significantly different between the above-mentioned groups: 42.3%, 10.8% and 0.0%, respectively (p = 0.005).

Conclusion: Hypofibrinolysis, physiologic fibrinolysis and hyperfibrinolysis can be differentiated as separate entities using ROTEM. Different rates of mortality are found across the fibrinolysis spectrum in the burn population. Hyperfibrinolysis is not a common expression of coagulation dysfunction in traumatic burn injury. Different distributions in the trauma and the burn population imply different mechanisms of dysfunction and limit generalizability of treatment standards across populations.      

102.09 Layperson Perception of and Ability to Apply an Improvised Tourniquet after B-Con Training

Posted on January 31, 2019

J. C. McCarty1,2, J. P. Herrera-Escobar1, Z. G. Hashmi1, M. A. Chaudhary1, E. De Jager1, C. J. Ezeibe1, R. M. Nunez1, A. H. Haider1, E. Goralnick1,3, E. J. Caterson1,2  1Brigham And Women’s Hospital,Center For Surgery And Public Health, Department Of Surgery,Boston, MA, USA 2Brigham And Women’s Hospital,Division Of Plastic Surgery,Boston, MA, USA 3Brigham And Women’s Hospital,Department Of Emergency Medicine,Boston, MA, USA

Introduction: The American College of Surgeons Bleeding Control Basic (B-Con) course is the most common hemorrhage control training for laypeople; teaching participants skills on how to pack a wound, apply pressure, and apply a commercial tourniquet. In most scenarios in the civilian sector, however, a tourniquet would not be immediately available in the event of a trauma. The Hartford Consensus states improvised tourniquets are an option if a commercial tourniquet is not available, but with minimal supporting data. The objective of this study was to evaluate laypeople’s 1) ability to improvise a tourniquet after B-Con training and 2) evaluate what participant’s perceived actions before and after the training if a commercial tourniquet were not available.

Methods: B-Con course participants were evaluated on their ability to fashion and apply an improvised tourniquet to a high-fidelity Hapmed trainer, which simulates bleeding and provides an estimated blood loss (EBL), immediately after B-Con training. Participants were provided gauze, shoestring, a belt, and a rod to act as a windlass. No feedback was given to participants about which materials to use. Participants were administered questionnaires before and after the B-Con course, but before testing, assessing what participants would do if presented with life-threatening extremity bleeding in the absence of a commercial tourniquet. Descriptive statistics were used to describe the primary and secondary outcomes.

Results: 61 laypeople were evaluated. 32.8% (n=20) participants correctly fashioned and applied an improvised tourniquet. Of the available materials, 82.0% (n=50) used the windlass, 62.3% (n=38) used the shoelace, 47.5% (n=29) used gauze, and 18.0% (n=11) used the belt.  The leather belt broke in 45.5% (n=5/11) of cases. 11 participants did not use a windlass and had a 0% success rate. When a commercial tourniquet was not available, pre-training 27.9% would apply an improvised tourniquet and 72.1%(n=44) would apply pressure. Post-training, 26.2% (n=16) would apply an improvised tourniquet and 72.1% (n=44) would apply pressure. Of those that would place an improvised tourniquet post-training, 8 (50%) applied the tourniquet correctly. 66.7% (n=40) reported the tourniquet was the most important skill taught in the course and 23.3% (n=14) thought it was how to apply pressure with your hands.

Conclusion: Civilian laypeople are unlikely to have a tourniquet when called upon to respond to a bleeding victim and, even with ideal supplies, can improvise a tourniquet less than a third of the time. The emphasis on tourniquet training for laypeople, rather than teaching pressure and packing alone, should be re-evaluated to align with the scenarios laypeople are likely to face.

102.08 Morning Report Decreases Length of Stay in Emergency General Surgery Patients

Posted on January 31, 2019

J. R. Gardner1, J. D. Wolfe1, W. Beck1, A. Bhavaraju1, M. K. Kimbrough1, B. Davis1, A. Privratsky1, M. Jupin1, J. Jensen1, R. Roberston1, K. Sexton1, J. R. Taylor1  1University of Arkansas for Medical Sciences,Department Of Surgery,Little Rock, AR, USA

Introduction:
Reduction of errors, as the result of inefficient patient hand-offs, has been a focus of interest in the Emergency General Surgery (EGS) field. High patient loads, and limited time to thoroughly conduct patient hand-offs during shift changes can result in errors in patient care. UAMS, a state-wide tertiary care center, changed its hand-off method from an email- based approach to a Morning Report (MR) model.

Methods:
Prior to MR, there was a lack of attending physician supervision during hand-offs between resident teams. The MR model instituted meetings between resident teams and 3 attending surgeons (night call, trauma day call, and EGS day call). The enterprise data warehouse was queried for all patients admitted to the Emergency General Surgery service from May 2014 until January 2018. Bivariate frequency statistics and linear regression analysis were performed using JMP Pro Version 13.2.1. Elixhauser categories were used for risk stratification.

Results:
2592 patients were analyzed in this study (pre-MR, n= 608; post-MR, n= 1984). The majority of patients were white males. The pre-MR cohort had an average age of 47.8 years compared to the post-MR cohort, 49.6 years (p= 0.253). 1484 patients had an operation, while 1108 did not. No significant difference in the number of comorbidities was found between pre and post-MR cohorts (p= 0.686). The LOS (days) for pre and post cohorts were (4.87 ± 7.9 and 4.13 ± 6.7, p= 0.019). Linear Regression showed procedures, Elixhauser Categories, Morning Report, age > 65, race, and gender were predictive of LOS.

Conclusion:
Attending supervised EGS MR is associated with a decreases length of stay. Further work needs to be done to quantify the effects of a MR system.
 

102.07 What’s Behind the Widened Mediastinum on CXR?

Posted on January 31, 2019

G. Vasileiou1, S. Qian1, H. Al-ghamdi1, D. Pace1, R. Rattan1, G. D. Pust1, M. Mulder1, N. Namias1, D. D. Yeh1  1University Of Miami,Surgery,Miami, FL, USA

Introduction:  It is commonly taught that a widened mediastinum (WM) on CXR is a marker for aortic injury (AI). We sought to describe the epidemiology of injuries for all patients with WM and compare their CXR to those of patients with confirmed AI. 

Methods:  Adults (age≥ 18) sustaining blunt traumatic injuries from 1/17-6/17 with both CXR (supine, anterior-posterior [AP]) and chest CT were included. We excluded those whose CT preceded CXR and those with missing data. Basic demographic information, injury characteristics, mediastinal width (MW), mediastinal-to-thoracic width ratio (MTR), and all thoracic imaging findings were analyzed. MW >8cm was considered “widened”.  We also queried our registry for all AI patients over a 4 yr period. The sensitivity (Sn), specificity (Sp), positive predictive value (PPV), negative predictive value (NPV), and accuracy of WM on CXR for AI were calculated for the 6-month period. Mann-Whitney U test was used as appropriate to compare patients with WM, and patients with confirmed AI. Multivariate logistic regression was performed to identify factors associated with positive traumatic findings. 

Results: Of 749 included subjects, 502 (67%) had a MW > 8 cm: mean age was 48 ±20 yrs, 381 (76%) were men, and mean BMI was 28 ± 5 kg/m2. Mechanism of injury was: motor vehicle crash (MVC) in 335 (67%); fall in 113 (23%); assault in 31 (6%), other (jet-ski accidents, etc.) in 17 (3%), and unknown in 6 (1%). Only 128 (26%) of WM patients had positive findings on CT, with the most common [80 (16%)] being non-traumatic findings (thymic tissue, lymph nodes, etc.), followed by hemo/pneumomediastinum [32 (6%)], sternal fractures [18 (4%)], multiple findings [15 (3%)], and vertebral fractures [6 (1%)]. Only 2 (1%) had AI. The Sn was 100%, Sp was 33%, PPV was 0.4%, NPV was 100%, and accuracy was 33%. From 2013-2017, 38 patients had AI: mean age was 46 ± 19 yrs, 26 (68%) were men, and mean BMI was 28 ±4 kg/m2. MVC was the most common mechanism (n=34 (89%)), followed by ‘other’ trauma mechanism in 2 (5%), fall in 1 (3%), and assault in 1 (3%). On univariate analysis, compared to all patients with WM, AI patients had significantly greater MW (9.5 [8.8-10.4] vs 10.2 [9.1-11.1]; p= 0.042) and MTR (0.31 [0.28-0.34] vs 0.32 [0.31-0.37]; p=0.001), though the actual differences were not clinically significant. Regression analysis did not identify any factors associated with traumatic CXR findings (Table).

Conclusion: Most blunt mechanism injured adults have a WM and the majority of those have either no findings or non-traumatic findings. The PPV of a WM for AI is <1%.  WM on supine AP CXR is non-specific and inaccurate for diagnosing traumatic injuries, especially AI.

 

102.06 Vices-Paradox in Trauma: Positive Alcohol and Drug Screens Associated with Decreased Mortality

Posted on January 31, 2019

J. Covarrubias1, A. Grigorian1, J. Nahmias1, T. Chin1, S. Schubl1, V. Joe1, M. Lekawa1  1University Of California – Irvine,Department Of Surgery,Orange, CA, USA

Introduction: There is a previously established association between trauma and alcohol, illegal drugs, as well as prescription drugs, all of which can lead to impaired judgement and reaction time resulting in injury. Improved survival in trauma patients with acute alcohol intoxication has been previously reported. The effect of illegal and prescription drugs on mortality is less clear. We hypothesized that alcohol, illegal and prescription drugs are each independently associated with decreased risk of mortality in adult trauma patients.

Methods: The Trauma Quality Improvement Program (2010-2016) was queried for patients screening positive for alcohol, illegal or prescription drugs on admission. These do not include prescription drugs for medical treatment. A multivariable logistic regression model was used to determine risk of mortality.

Results: From 1,299,705 adult patients, 227,995 (17.5%) screened positive for alcohol, 155,437 (12.0%) for illegal drugs and 90,259 (6.9%) for prescription drugs. The alcohol cohort had the highest mortality rate (6.2%), followed by prescription drugs (5.7%) and illegal drugs (5.1%) (p<0.001). After controlling for covariates in an analysis of all adult trauma patients, all three groups had lower risk for mortality: alcohol (OR=0.88, CI=0.84-0.92, p<0.001), illegal drugs (OR=0.80, CI=0.74-0.86, p<0.001), prescription drugs (OR=0.70, CI=0.65-0.76, p<0.001). When stratified by injury severity score (ISS), those screening positive for alcohol or illegal drugs continued to have decreased mortality until an ISS of 50. Patients screening positive for prescription drugs were associated with decreased mortality when ISS>16.

Conclusion: Compared to all trauma patients, those screening positive for alcohol on admission have more than a 10% decreased risk of mortality, those screening positive for illegal drugs have a 20% decreased risk of mortality, and those screening positive for prescription drugs have a 30% decreased risk of mortality. The effect of alcohol and illegal drugs on risk for mortality ceases only when ISS>50. This paradoxical association should be confirmed with future clinical studies, as well as merits basic science research to help identify biochemical or physiologic components conferring a protective effect on survival in trauma patients.

102.05 Fat Embolism – a Serious Complication after Trauma: An Analysis of the National Trauma Data Bank

Posted on January 31, 2019

G. Vasileiou1, J. Parks1, D. D. Yeh1, R. Rattan1, T. Zakrison1, N. Namias1, G. D. Pust1  1University Of Miami,Surgery,Miami, FL, USA

Introduction: Fat embolism (FE) is a rare complication after trauma that may have devastating consequences. Our objective was to describe the incidence of and clinical outcomes after FE.

Methods:  The 2008- 2014 National Trauma Data Bank (NTDB) were queried for FE, using ICD9 958.1. Descriptive analysis of demographics, injury, and hospitalization characteristics was carried out. Multivariate logistic regression analysis for mortality was performed controlling for age, gender, ISS score, and intensive care unit (ICU) admission. 

Results: Out of 4,495,935 patients, we identified 418 (0.01%) diagnosed with FE. Median age was 32 [21-62], and 275 (66%) were male. Of those, 393 (94%) had blunt mechanism, and 407 (97%) had fractures; femur fractures were the most common 286 (68%); followed by multiple bone fractures 225 (54%); other (ribs, vertebra, radius, ulna, etc) bone fractures 200 (48%); tibia and/or fibula 130 (31%); pelvis 78 (19%); and humerus 31 (7%) fractures. Median time to OR was 15 [6 – 34] hours.Internal fixation was performed in 248 (59%), followed by procedures without internal fixation in 47 (12%), and removal of implants in 35 (9%) patients. ICU admission was required in 291 (70%) patients, and 152 (36%) needed mechanical ventilation for 6 [2-11] days. Median hospital length of stay (LOS) and ICU LOS were 10 [6-16] and 5 [2-12] days, respectively. Median Injury Severity Score (ISS) was 10 [9-18], yet in-hospital mortality was 13% (n=54). There were 169 (40%) patients that were discharged home with or without additional services while 184 (44%) were transferred to other facilities. Multivariate logistic regression analysis showed that mortality was associated with ventilation (OR: 4.05; 95% CI [2.01-8.13]; p<0.001), age (OR: 1.02; 95% CI [1.01-1.03]; p=0.006), and ISS (OR: 1.04; 95% CI [1.01-1.07]; p=0.018), (Table 1).

Conclusion: Fat embolism is an extremely rare complication that occurs almost exclusively in trauma patients with fractures.  FE is associated with higher-than-expected mortality based on ISS and most patients require ICU admission and usually with concomitant mechanical ventilation. Formal diagnostic criteria and severity grading is the next step required for improving diagnostic accuracy and treatment of this entity.

 

102.04 The Impact of Obesity on Severity and Outcomes in Penetrating Abdominal Trauma

Posted on January 31, 2019

E. De La Cruz1, O. A. Olufajo1, A. Zeineddin1, E. Cornwell1  1Howard University College Of Medicine,Surgery,Washington, DC, USA

Introduction:
Obesity is widely acknowledged to be a predictor of increased morbidity and mortality. Multiple studies investigating the association between body mass index (BMI) and blunt abdominal trauma have shown a protective effect of increasing BMI on the severity of injury presumably owing to a “cushion effect”. However, the number of studies exploring the association of BMI and the severity of abdominal penetrating trauma patients is rather limited. The aim of our study is to evaluate that association using a nationwide sample.

Methods:
Data was retrieved from the National Trauma Data Bank (2013-2015). Patients included were those with penetrating abdominal trauma. Patients were stratified by BMI status (<18.5, 18.5-24.9, 25-29.9, 30-39.9, >40). Patients without information on BMI were excluded from the analyses. We defined injury severity using two methods. First, we used the abdomen abbreviated injury scale (AIS) ranging from 1 to 6. Second, we categorized patients as having an abdominal operation vs. no abdominal operation. We evaluated differences in injury severity and mortality across BMI groups using Chi-square tests. Logistic regression multivariate regression models were used to identify independent associations between BMI and the outcomes measured.    

Results:
We included 22,110 patients with abdominal penetrating trauma: 10,856 stab wounds, (SW) and 11,254 gunshot wounds (GSW). With increasing BMI, there was a decrease in AIS>2 in SW (26.4%, 27.2%, 26.9%, 23.1%, 20.9%) (P<0.001) and in GSW (60.4%, 51.7%, 52.7%, 50.4%, 48.1%) (P=0.016). The rate of operative management across BMI groups in SW was 43.6%, 43.7%, 43.3%, 44.8%, 46.1% (P<0.655), and in GSW was 59.2%, 58.9%, 59.8%, 60.9%, 54.9% (P=0.084). On multivariate analysis, patients with BMI 30-39.9 had increased odds of undergoing surgical procedures compared to patients with normal BMI [Odds Ratio, OR (95% Confidence Interval, CI)]: 1.15 (1.01-1.30) among those with GSW, but there was no difference in SW. The unadjusted analysis showed an increase in mortality with increasing BMI among patients with GSW (5.8%, 5.9%, 5,2%, 6.9%, 7.8%) (P=0.024), but no difference in mortality with increasing BMI in SW (1.4%, 1.3%, 1.2%, 0.9%, 2.5%) (P= 0.096). However, on multivariate analysis of patients with SW, patients with BMI >40 had increased odds of mortality compared with patients with normal BMI [OR(95% CI): 2.35 (1.08 – 5.06)]. This was also true for patients with BMI >40 among patients with GSW [OR(95% CI): 1.89 (1.26 – 2.86)].

Conclusion:
Increased BMI seems to have a protective effect against penetrating abdominal trauma, as it was associated with lower incidence of severe injury. However, there is increased mortality in morbidly obese patients who were victims of penetrating abdominal trauma. This study suggests that the protective effect of obesity in injury severity may be countered by other factors inherent to the morbidly obese population.

102.03 Regional Survey of Chest Tube Management Practices by Trauma Surgeons

Posted on January 31, 2019

M. H. Parker1, A. Newcomb1, C. Liu1, C. Michetti1  1Inova Fairfax Hospital,Falls Church, VA, USA

Introduction:
Evidence to guide CT management in trauma patients is limited and tends toward thoracic surgery patients. The goal of this study was to identify current practices among trauma providers regarding trauma CT management in trauma patients.

Methods:
We designed a web-based survey (Survey Monkey) to assess CT management practices of trauma providers who were active, senior, or provisional members (N=1890) of the Eastern Association for the Surgery of Trauma via email. The survey contained multiple choice and write-in questions. Descriptive statistics were used.

Results:

The response rate was 39% (N=734). 91% of respondents were attending surgeons, the remainder fellows or residents. Attendings were more likely than trainees to place pigtail catheters for stable patients with pneumothorax (PTX). Attendings with <5 years’ experience were more likely to choose a pigtail than more experienced surgeons for elderly patients with PTX. Respondents preferred standard size CT for hemothorax (HTX) and unstable patients with PTX, and larger tubes for unstable patients with HTX (Figure 1).

97.3% (PTX) and 97.5% (HTX) would place a CT to suction following placement.  Most respondents (58.9%) selected transitioning to water seal after resolution of any air leak, but not before 24 hours.  25.7% would use water seal after resolution of air leak regardless of timing.  For hemothorax, 41.9% of respondents would place to water seal based on a specific fluid output, 27.7% after 24 hours and 19% based on CXR findings.  While CT was on suction, the majority of respondents would allow water seal for ambulation for PTX (85.0%) and HTX (93.4%).  The median output at which respondents would remove a chest tube was 150cc for serosanguinous fluid and 100cc for bloody fluid.  After CT removal for PTX, CXR was preferred at 4 hours (39.7%), 6 hours (21.0%) 1 hour (13.8%); 12.9% did not get a CXR.

For non-ventilated patients, most attendings chose to get CXR after placement (96.7%), prior to removal at the end of a water seal trial (69.4%) and after removal (66.1%).  Some preferred CXR prior to placement to water seal (45.3%) or daily CXR (38.9%).  At outpatient follow-up, only 27.4% would get a CXR for PTX.  The majority (53%) perceived the quality of evidence for trauma CT management to be low and cited personal experience and training as the main factors driving their practice.

Conclusion:

Trauma CT management is variable and non-standardized, and depends mostly on clinician training and personal experience.  Few surgeons identify their practice as evidence based. We offer compelling justification for the need for trauma CT management research to determine best practices.

102.02 The Revolving Door, Readmissions after Traumatic Brain Injury

Posted on January 31, 2019

A. Brito1, L. N. Godat2, A. E. Berndtson2, J. Doucet2, A. M. Smith2, T. W. Costantini2  1University Of California – San Diego,General Surgery,San Diego, CA, USA, 2University Of California – San Diego,Division Of Trauma, Surgical Critical Care, Burns And Acute Care Surgery – Department Of Surgery,San Diego, CA, USA

Introduction: Traumatic brain injury (TBI) is associated with functional deficits, impaired cognition and medical comorbidities that continue well after the initial injury.  Many patients seek medical care at other healthcare facilities following discharge, rather than returning to the admitting trauma center, making assessment of readmission rates difficult to determine.  The objective of this study was to determine the incidence and factors associated with readmission to any acute care hospital after an index admission for TBI.

Methods: The Nationwide Readmission Database was queried for all patients admitted with a TBI during the first 3 months of 2015.  Readmissions for this population were then collected for the remainder of 2015.  Patients that died during the index admission were excluded. Demographic data, injury mechanism, type of TBI, the number of readmissions, days from discharge to readmission, readmission diagnosis and mortality were studied.

Results: Of the 15,277 patients with an index admission for TBI, 5,296 patients (35%) required at least 1 readmission. The number of readmissions ranged from 1 to 14.  Twenty six percent of readmissions occurred within the first 2 weeks after discharge from the index trauma admission (see Figure).  Patients with subdural hematoma (SDH) were more likely to require readmission compared to other types of TBI (RR 1.21, p<0.001).  The most common primary diagnosis on readmission was SDH, followed by septicemia, urinary tract infection, and aspiration.  The 3 most frequent injury mechanisms associated with readmission were fall (86%), motor vehicle crashes (7%) and assaults (4%).  Readmission rates increased with age, with 94% occurring in patients over the age of 45 and 75% in patients >65 years.  Mortality ranged from 6-14% on depending on the number of subsequent readmissions after TBI.

Conclusion: Hospital readmission is common for patients discharged following TBI.  Elderly patients who fall with resultant SDH are at especially high risk for complications and readmission. Understanding potentially preventable causes for readmission can be used to guide discharge planning pathways to decrease morbidity in this patient population.

102.01 Development and Generalization of a Score to Predict Trauma Patient Discharge Disposition using NTDB

Posted on January 31, 2019

M. Graham1, P. Parikh1,2, S. Hirpara2, M. McCarthy1, P. P. Parikh1  1Wright State University,Department Of Surgery,Dayton, OH, USA 2Wright State University,Department Of Biomedical, Industrial, And Human Factors Engineering,Dayton, OH, USA

Introduction: Delay in discharge planning could result in extended length of stay leading to increased hospital costs, ineffective utilization of resources, and delays in rehabilitation treatment in trauma patients. Limited work has been done in developing models predicting discharge disposition in trauma patients. These models are developed using a single institution data and have not be demonstrated to be generalizable. The objective of this study is to develop a predictive model using the National Trauma Data Bank (NTDB) and evaluate its generalizability on data from a Level I trauma center.  

Methods:  NTDB data from 2015 were used to build and validate a binary logistic regression model using derivation-validation (i.e., train-test) approach to predict patient disposition location (home vs nonhome) upon admission. Patient demographics and clinical variables available at the time of admission were considered in the analysis. A Mann-Whiney U-test was used to compare patient parameters. The regression model was then converted into a 20-point score using an optimization-based approach. An appropriate threshold was selected to achieve a score with a sensitivity of >0.80 and specificity of >0.50. The generalizability of this score was then evaluated on the trauma registry data at our Level I trauma center in Midwest US.

Results:A total of 558,599 cases in the NTDB were included in the study, out of which, 178,666 (31.98%) went to a nonhome location and 379,933 (68.02%) patients went home. The average age of patients with a nonhome disposition compared to home disposition was significantly higher (68.11 ± 20.69 years vs. 43.23 ± 23.09 years; p<0.001) and had more severe injuries measured using the ISS (11.26 ± 8.25 vs 8.04 ± 6.28; p<0.001). Increased age, female sex, higher ISS, and the comorbidities of cancer, cardiovascular, coagulopathy, hepatic, neurological, psychiatric, renal, substance abuse, and diabetes were independent predictors of nonhome discharge. The logistic regression model’s AUC was 0.83; the score achieved a correlation of 0.94 with the predicted probabilities from the regression model. A threshold value of 4 or higher indicated higher likelihood of nonhome discharge; this threshold resulted in a sensitivity of 0.86 and specificity of 0.62 on NTDB validation data (n=167,580). The score generalized well on the insitutional data (n=3,384) obtained from trauma registry of our Level I Trauma Center; sensitivity of 0.85 and specificity of 0.60.

Conclusion:A model and a score developed using NTDB could be implemented at a Level I trauma center to predict upon admission a trauma patient’s discharge disposition location, home or nonhome. This score can aid in early hospital preparation for patients predicted to be discharged to a nonhome location yielding a smoother transition, increased satisfaction, effective utilization of hospital resources, and potentially decrease total operating costs.

100.18 Hospital Readmission Following Discharge from the NICU–A Pediatric Health Information System Study

Posted on January 31, 2019

M. Joseph1, M. A. Bartz-Kurycki1, J. K. Chica1, K. Tsao1, M. T. Austin1  1McGovern Medical School at UTHealth,Pediatric Surgery,Houston, TX, USA

Introduction: Approximately 8% of all live births in the US require admission to the neonatal intensive care unit (NICU) after birth.  Prior studies have shown that these infants are at significantly increased risk of readmission after discharge with readmission rates ranging from 15-50%.  Our prior work identified minority race/ethnicity as the strongest predictor of readmission following NICU discharge in our hospital system.  In this study, we aimed to determine the incidence of 90 day hospital readmission for infants discharged from the NICU and determine patient characteristics that increase likelihood of readmission using a large national database.

Methods: The Pediatric Health Information System (PHIS) nationwide was queried from 2016-2018 for patients discharged from the NICU. Deceased patients and those discharged within three days of admission were excluded. Descriptive statistics and univariate and multivariate logistic regression were tabulated utilizing SPSS (v24) to determine factors associated with readmission to the hospital within 90 days of discharge from the NICU.

Results: 86,114 patients were included in the final sample. The majority were non-Hispanic white (NHW) (49.6%) followed by non-Hispanic black (NHB) (15.5%), and Hispanic (14.5%). Most were publically insured (n=51,800, 56%).  The median gestational age was 35 weeks (IQ range 26-38) and median birthweight was 2780gm (IQ range 1910-3374gm).  53,914 (63%) were readmitted to the hospital within 90 days of discharge from the NICU.  NHB race/ethnicity and the diagnosis of a complex chronic medical condition were associated with increased odds of readmission (OR 1.04, 95% CI 1.01-1.08 and OR 1.04, 95% CI 1.01-1.07, respectively).  After controlling for other demographic and clinical factors, only complex chronic medical condition was associated with 90-day readmission (OR 1.05, 95% CI 1.02-1.08).

Conclusions: Infants are readmitted at an alarmingly high rate following discharge to home from the NICU; however, few factors were identified to be associated with readmissions using this dataset.  Race/ethnicity may play a role but the causes of readmission are likely multifactorial.  These will only be addressed through future study that uses both quantitative and qualitative methods to identify potential modifiable risk factors for readmission in this high risk population.

09.19 Primary Hyperparathyroidism and Bone Density in Patients with a History of Roux-en-Y Gastric Bypass

Posted on January 31, 2019

V. Lai1, K. D. Burman2  1Virginia Hospital Center,Arlington, VA, USA 2MedStar Washington Hospital Center,Washington DC, WASHINGTON, DC, USA

Introduction:  Primary hyperparathyroidism (1HPT) affects 1-3% of the population and can negatively impact bone mineral density (BMD), with an increased risk of osteoporosis and fractures.  A much higher percentage of the population has obesity, and rates of Roux-en-Y gastric bypass (RYGB) surgery to correct morbid obesity has increased.  Some have noted that RYGB patients may develop lower BMD, especially at the femur.  Secondary hyperparathyroidism (2HPT) in RYGB patients is common, but 1HPT in RYGB patients has not been well studied, particularly studies of their bone health.  The aim of the study was to compare the BMD of RYGB patients who develop 1HPT to those with 1HPT who have not undergone RYGB.  The hypothesis was that patients with 1HPT and a history of a RYGB would have lower BMD than controls.

Methods: A retrospective review of adult patients with sporadic 1HPT cared for within a multi-site metropolitan health network between 2000-2018 was performed.  Patients with a history of RYGB and 1HPT were identified with ICD and CPT codes, and included if they had BMD data from dual-energy x-ray absorptiometry scans. Cases were matched 1:1 by age, race, and sex to a control group of patients with 1HPT without a history of RYGB. BMD, biochemical, and clinical data were collected.  

Results: Four patients with a history of RYGB who developed 1HPT were identified: 100% were female; 50% were white and 50% were black; the average age at the time of 1HPT diagnosis was 61 years.  The cohort was more likely than the controls to have osteoporosis (75% vs. 25%) and less likely to have any one site with normal BMD (0% vs. 100%).  The worst BMD occurred in the distal radius and lumbar spine in the cohort group, and in the lumbar spine in the control group.  Fractures occurred in 50% of both.  The patients with a history of RYGB with 1HPT tended to have lower serum calcium and higher parathyroid hormone (PTH).  Both groups were vitamin D replete, and the RYGB group was more likely to have taken high-dose supplementation (75% vs. 25%) to achieve vitamin D repletion. All patients in the cohort group underwent parathyroidectomy without significant complications, and with postoperative normalization of serum calcium and PTH.

Conclusion: Patients with 1HPT who have undergone RYGB may present with worse BMD than those with 1HPT who have not undergone RYGB.  The distribution of the bone disease in the patients with 1HPT and a history of RYGB seemed more similar to the pattern of bone disease of typical 1HPT patients than in typical post-RYGB patients, where changes occur in the femur.  They were more likely to have required high-dose supplementation to be vitamin D replete, and they had lower serum calcium and higher PTH levels, which may reflect the influence of 2HPT to the 1HPT picture.  Whether this contributed to the BMD results is possible, and further study on the bone health of these patients would help clarify the results of these initial findings.
 

09.16 Is Breast Cancer Associated with Primary Hyperparathyroidism?

Posted on January 31, 2019

R. Arrangoiz1, D. Margain1, D. Margain1, J. Sanchez1, F. Cordera1, D. Caba1, M. Muñoz1, E. Moreno1, E. Luque1, R. Arrangoiz1  1Sociedad Quirúrgica S.C. at the American British Cowdray Medical Center,Surgical Oncology / Head And Neck Surgery / Endocrine Surgery,Mexico City, MEXICO CITY, Mexico

Introduction: Breast cancer is the leading malignancy in women and the second leading cause of cancer-related deaths across the world. Hypercalcemia is known to occur in up to 20% to 30% of the patients with cancer at some point during the course of their illness. Breast cancer is one of the malignancies most commonly associated with hypercalcemia. Primary hyperparathyroidism (PHPT) has been associated with an increased risk of developing breast cancer compared with patients without PHTP.  Little is known about the underlying risk factors. The aim of our study is to describe a cohort of patients with PHPT and breast cancer.

 

Methods: Retrospective study from a prospectively kept database of patients with PHPT treated by our group between January 2015 and July 2017 who had been diagnosed with breast cancer. The patients’ characteristics were obtained and analysed from the electronic medical records. Patients without complete medical records were not included in our study. All data were collected in a non-identifiable fashion in accordance with the principles outlined in the Declaration of Helsinki and as required for our institutional review board approval.

 

RESULTS AND DISCUSSION: A total of ten patients from a database of 75 patients were included in this study. All patients were female, the mean age was 59.2 years. The mean preoperative calcium, PTH and vitamin D was 10.1 mg/dL, 99.6 pg/mL and 25.5 ng/dL, respectively. Significant decreased of intraoperative PTH and postoperative calcium and PTH were achieved after surgical treatment. Pathology reported that 50% of the cases were secondary to a single adenoma (five patients) and 50% (five patients) of cases with hyperplastic parathyroid glands. Unilateral (70%), stages I or II (70%), invasive ductal breast carcinoma (90%) was the most common diagnosis. The immunohistochemical status reported that 80% of patients had positive hormone receptors. The mean time between breast cancer and PHPT operations was 89.5 months.

 

Conclusion: Breast cancer and PHPT share several common characteristics, which has led to the postulation of common etiological pathways. However, the exact pathogenesis and the relationship between breast cancers and PHPT still remains obscure. PHTP should be considered as a possible cause of hypercalcemia in patients with non-aggressive breast cancer. We suggest that serum PTH should be determined in all breast cancer patients with increased serum calcium concentration, especially in those with no evidence of metastatic disease. 

 

07.19 The 100 Most Influential Manuscripts in Robotic Surgery: A Bibliometric Analysis

Posted on January 31, 2019

T. M. Connelly1, Z. Malik1, R. Sehgal1, J. C. Coffey1, G. Byrnes1, C. B. Peirce1  1University Hospital Limerick,Colorectal Surgery,Limerick, CO. LIMERICK, Israel

Introduction:  Since the first robotic assisted surgery in 1985, a steady increase in the number of procedures performed annually has been documented.  Bibliometric analysis highlights the key studies that have influenced current practice in a field of interest.  We use bibliometric analysis to evaluate the 100 most cited manuscripts on robotic surgery and discuss their influence on the evolution of the platform.

Methods:  The terms ‘robotic surgery,’ ‘robot assisted surgery’ and ‘robot-assisted surgery’ were used to search Thomson Reuters Web of Science database for full length, English language manuscripts. The top 100 cited manuscripts were analyzed by manuscript type, surgical specialty, first and last author, institution, year and journal of publication

Results: 14,980 manuscripts were returned. Within the top 100 cited manuscripts, the majority featured urological surgery (n=28), followed by combined results from surgical subspecialites (n=15) and colorectal surgery (n=13). The most cited paper authored by Nelson et al (432 citations) reviewed technological advances in the field. European Urology published the highest number of papers (n=15,  2595 citations). The year and country with the greatest number of publications were 2009 (n=15) and the USA (n=68).  The Johns Hopkins University published the most top 100 manuscripts (n=18).

Conclusion: The 100 most cited manuscripts describe the progression of robotic surgery from a basic instrument-holding platform to today’s articulated instruments with 3D technology. From feasibility studies to multicenter trials, this analysis demonstrates how robotic assisted surgery has gained acceptance in urological, colorectal, general, cardiothoracic, orthopedic, maxillofacial and neurosurgery.

 

07.17 Carcinoma of Unknown Primary: Incidence and Clinical Course

Posted on January 31, 2019

T. M. Connelly1, M. Alzamzami1, M. Foley2, M. S. Khan1, A. Mehmood1, G. T. O’Donoghue1  1University Hospital Waterford,General Surgery,Waterford, WATERFORD, Ireland 2Waterford Institute of Technology,School Of Health Sciences,Waterford, WATERFORD, Ireland

Introduction:  Carcinoma of unknown primary (CUP) is an unpredictable and difficult diagnosis in up to 3% of cancer patients. We aimed to determine the incidence and clinical course of patients diagnosed with cancer CUP in the South East Ireland Regional Cancer Center.

Methods:  Gastrointestinal multidisciplinary meeting (MDM) records of 4274 patients were used to identify those initially diagnosed with CUP between September 2008-December 2017.  Patient demographics, radiology, endoscopy, pathology, hematology and biochemistry blood results were analyzed. 
 

Results: 97 patients were initially diagnosed with CUP.  In 50 (51.5%, Table 1), a primary was never determined by the study end. In this cohort (50% male, mean age=63.4 years), 94% underwent ≥one biopsy.  The most common histological subtypes were adenocarcinoma (59.1%) and neuroendocrine tumors (18.2%).   When comparing the CUP cohort to the cohort in whom a primary was determined (n=47, 48.5%), there was no difference in gender distribution, mean age, referral source or number of times discussed at MDM, biopsies or surgical interventions. CUP patients underwent twice as many radiological investigations (median 6 vs 12,p=0.003) and had a higher median CEA (5.8 vs 2.6,p=0.02). At the study endpoint, 72% of those with a primary diagnosed vs 82% of CUP patients were deceased (p=0.09). 

Conclusion:  In approximately half of patients initially diagnosed with CUP in an MDM setting, a primary is not determined. Mean survival is less than 8 months. These patients undergo more radiological investigations when compared to those in whom a primary is determined. However, no difference in mean age, number of biopsies or surgical interventions is found.

 

07.15 Using Data Envelopment Analysis for Improving Allocation of Operating Room Capacity among Surgeons

Posted on January 31, 2019

V. Tiwari1, D. Penson1, W. Sandberg1  1Vanderbilt University Medical Center,Nashville, TN, USA

Introduction: In Academic Medical Centers, surgical specialties are allocated operating room (OR) capacity, which they in turn split among individual surgeons. Surgeons thus get the coveted access to first-case placement priority in their OR(s). This bi-level capacity disaggregation gives surgeons certainty and control over their schedules; however, it restricts access of those ORs to other surgeons, often leading to underutilization. It also makes meaningful interpretation of service-level OR utilization challenging. We developed a method to rank surgeons’ based on their use of OR time. We use the ranking to recommend to service chiefs which surgeons need more or less OR capacity.

Methods: Data Envelopment Analysis (DEA) is a non-parametric mathematical optimization technique to rank peers based on their technical efficiency in converting multiple inputs (or resources) into multiple outputs (or performance measures). DEA assigns a score between one (most efficient individual) and zero (those deemed least DEA efficient). We developed four input variables (total unique days in OR, number of days starting one OR with first-case, as well as with two ORs, and number of total first-starts), which each reflect in some manner the amount of capacity consumed by a surgeon. We developed four output measures that each reflect how well the ORs were used each day that the surgeon started OR(s) with a first case. We used 1 year of data and repeated the analysis four times, once for each quarter, to rank each of the 175 surgeons. Analysis was conducted using Microsoft MS-Excel and the open source DEA software OSDEA.

Results: Surgeons’ rank between quarters remained consistent, implying stable practice patterns, at least in the short-term. DEA finds those surgeons more technically efficient that run their ORs to much later in the day (that is, higher utilization); even if they have lower overall surgical case volumes, but consume relatively fewer resources (that is, fewer first-starts). In addition, the surgeons that start two ORs with first-cases tended to have lower DEA efficiency scores. This is because these surgeons usually had high OR utilization in just one of their ORs and finished the other OR around noon, while not producing substantially higher overall surgical case volume as compared to those surgeons that just started one OR but used it well. The method also helped identify those (usually younger faculty) surgeons that do not have an assigned OR (lower resource consumption), but often were able to get a first-case start through the released-room policy. This DEA technique thus aids in the natural selection of surgeons that should get (or lose) OR time.

Conclusion: We developed a method that objectively assists surgical service chiefs in reallocating capacity among surgeons. The technique compares an individual’s performance relative to their peers with respect to how efficiently the individual converts a set of inputs into outputs.

05.20 Characteristics of Breast Cancer in Young Patients Based on a Saudi Tertiary Hospital Experience

Posted on January 31, 2019

A. Alhefdhi1, S. AlNefaie1, O. Almalik1  1King Faisal Specialist Hospital and Research Center (KFSH), and AlFaisal University,Breast And Endocrine Surgery/General Surgery,Riyadh, RIYADH, Saudi Arabia

Introduction: Breast cancer accounts for 26% of the cancers in Saudi Arabia, with a median age of 49 years at diagnosis. However, no published literature about treating breast cancer in a young Saudi population.The aim of this study is to explore the characteristic features of breast cancer in Saudi patients 40 years of age or younger.

Methods: A five-year retrospective review conducted, including all patients diagnosed with breast cancer and operated at a single tertiary hospital.

Results:A total of 1026 were identified. Among them, 230 cases (22%) meet our inclusion criteria with a mean age of 34±5 years. The majority had IDC (85.7%), followed by DCIS in (6.5%), malignant phylloids in (3.1%),  ILC in (1.7%), metaplastic carcinoma in (1.3%), intracystic papillary carcinoma in (0.9%), (0.4%) of borderline phylloid, and  (0.4%) of Angiosarcoma. Only (14%) found to have a positive family history of breast cancer, and (15%) found to have metastatic disease at diagnosis. The recurrence rate was (17.4%). Moreover, the majority underwent lumpectomy with/without SLND/ALND (38%), followed by SSM (22%), simple mastectomy with/without SLNB (13%), MRM (10%), palliative surgery (6.5%), nipple sparing mastectomy (4.5%), wire localization lumpectomy (4%),  then bilateral surgery (3%). Among the patients with IDC (51.3%) had grade 3 cancer, (70%) had a positive ER, (53%) had a positive PR,  (26%) had positive Her-2,  (24%) had triple negative.

Conclusion:Our data suggested that 1/4th of breast cancer cases are 40 years or younger. Further studies are needed to confirm our findings and to exclude possible genetic mutations.

 

05.17 Breast-Conserving Surgery for Lobular Carcinoma In Situ Variants: A Single Institution’s Experience

Posted on January 31, 2019

D. I. Hoffman1, P. J. Zhang2, J. Tchou1,3  1Perelman School of Medicine at the University of Pennsylvania,Department Of Surgery,Philadelphia, PA, USA 2Perelman School of Medicine at the University of Pennsylvania,Department Of Pathology,Philadelphia, PA, USA 3Perelman School of Medicine at the University of Pennsylvania,Abramson Cancer Center, Rena Rowan Breast Center,Philadelphia, PA, USA

Introduction: Lobular carcinoma in situ (LCIS) found on core needle biopsy is a benign lesion that confers increased lifetime risk of breast cancer but generally does not require further surgery. In contrast, non-classic LCIS (NC-LCIS), which includes high-grade variants with pleomorphism or necrosis, warrants surgical excision. In patients pursuing breast-conserving surgery (BCS) for NC-LCIS, the need for wide surgical margins to prevent recurrence is controversial. We therefore characterized the surgical management and outcomes of women diagnosed with NC-LCIS at a large, academic medical center.

Methods:  A retrospective database query was conducted to identify female patients seen at our institution from 2008–2018 with a biopsy diagnosis of NC-LCIS. Patients were excluded if NC-LCIS was diagnosed in the background of invasive breast carcinoma or ductal carcinoma in situ (DCIS). Clinicopathologic, surgical, and follow-up data were collected by chart review. Rates of upstage, re-excision, and recurrence were calculated.

Results: We identified 26 patients with NC-LCIS diagnosed on biopsy. The cohort was mostly white, the median age was 54 years (range 40–70), and half were postmenopausal. None were known carriers of breast cancer gene (BRCA) mutations, but 10 patients had a first-degree family history of breast cancer. Almost all (24/26) presented with an abnormal screening mammogram, 22 of which had suspicious calcifications. 80.8% (21/26) of patients initially pursued breast conservation, while 19.2% (5/26) underwent immediate mastectomy. At definitive surgery, 11.5% (3/26) were upstaged to DCIS or invasive carcinoma. Among 19 patients with a final diagnosis of NC-LCIS undergoing BCS, 47.4% (9/19) had at least one re-excision and five patients converted to completion mastectomy. In patients receiving BCS without completion mastectomy, 64.3% (9/14) had final surgical margins that were negative for NC-LCIS, while 35.7% (5/14) had close (<1mm) or positive margins. No recurrences in patients with negative margins at definitive surgery were observed. One patient with positive margins developed a local recurrence 8.3 years after surgery, and one patient with close margins did 2.2 years after surgery. All patients with a final diagnosis of NC-LCIS were alive at time of analysis with no evidence of progression to invasive carcinoma, mean follow-up time 4.5 years (range 20 days–10.5 years).

Conclusion: We presented the clinical outcomes of one of the largest single institution series of NC-LCIS, a rare diagnosis. In patients with a final diagnosis of NC-LCIS pursuing breast conservation, re-excisions are common and negative margins can be challenging. However, when negative margins are achieved, recurrence risk is low.

05.16 Frequency of PI3K pathway activation, CCNE1 amplification, Rb1 and ESR1 mutations in breast cancer

Posted on January 31, 2019

T. Takeshita1, E. Katsuta1, L. Yan2, K. Takabe1  2Roswell Park Cancer Institute,Biostatistics & Bioinformatics,Buffalo, NY, USA 1Roswell Park Cancer Institute,Breast Surgery, Surgical Oncology,Buffalo, NY, USA

Introduction:  Endocrine therapies are one of the essential treatments for estrogen receptor (ER)-positive breast cancer, particularly in combination with targeted therapies. It is now established that endocrine therapy with targeted therapies, such as, CDK4/6 inhibitors and mTOR inhibitors, is a mainstay of treatments for ER-positive metastatic breast cancer (MBC). ESR1 mutation is a resistance factor of endocrine therapy and it is a useful biomarker predicting an effect of the treatment, but it is not certain whether or not it has same utility when used with targeted therapies. RB1 mutation and amplification of cyclin E1 (CCNE1) are now known as resistance factors of CDK4/6 inhibitors. PI3K pathway is known to play an important role in ER positive breast cancer and its activity affects both endocrine therapy and molecular target therapy. Therefore, it is noteworthy whether coexistence of ESR1 mutation and these genetic abnormalities is a predictor of effect of these combination therapies. Here we studied the frequency of PI3K pathway alterations (PIK3CA mutation/amplification and PTEN loss), Rb1 mutation, and amplification of CCNE1 with ESR1 mutation in ER-positive breast cancer.

Methods: We analyzed gene abnormalities using an ER-positive primary breast cancer (PBC) cohort from TCGA data (n = 525) and a MBC cohort (n = 216).

Results: In the PBC cohort, PI3K pathway alterations were recognized at 40%, Rb1 mutation at 0.9%, amplification of CCNE1 at 10%, and ESR1 mutation at 0.9%. As expected, presence of ESR1 mutation was very low and it was difficult to verify the relationship with other factors. In the MBC cohort, the frequency of PI3K pathway alterations, Rb1 mutation, and amplification of CCNE1 was almost unchanged, but ESR1 mutation was found to be 20%. Some cases with genomic alterations coexisting with ESR1 mutation were found, suggesting the possibility of showing resistance to combination therapies.

Conclusion: We showed the clinically important frequency of genomic alterations coexisting with ESR1 mutation in ER positive breast cancer cohort.

« Previous Page
Next Page »