43.09 Etiologies and Disparities of Goiter in America

Posted on January 18, 2018

R. Zheng1, A. Rios-Diaz1, D. P. Thibault2, J. A. Crispo2, A. Willis2, A. I. Willis1  1Thomas Jefferson University,Department Of Surgery,Philadelphia, PA, USA 2University Of Pennsylvania,Department Of Biostatistics And Epidemiology,Philadelphia, PA, USA

Introduction:
Although iodine deficiency is a known cause of goiter worldwide, there is a paucity of information regarding people undergoing operations for simple, iodine-deficiency-related goiters in the U.S. Our study aimed to determine the current distribution of etiologies of goiters undergoing thyroidectomy in the U.S., and to identify disparities among those most at risk for simple goiters (SG).

Methods:
The National Inpatient Sample (NIS) database was queried to identify hospitalizations of individuals who underwent complete or partial thyroidectomy in years 2009-2013, using International Classification of Diseases, Ninth Revision, Clinical Modification codes to classify hospitalizations by goiter subtype. Descriptive statistics were used to analyze patient- and facility-level factors by chi square. Multivariable logistic regression models were used to determine factors associated with the development of SG requiring thyroidectomy.

Results:
We identified 103,678 hospitalizations for thyroidectomy: 58,313 total; 38,433, partial; 6,391, substernal. SG diagnosis was coded in 7,692 (7.42%); nodular goiter, 73,524 (70.92%); thyrotoxicosis, 14,043 (13.54%); thyroiditis, 1,248 (1.20%); and thyroid cancer, 7,169 (6.92%). Significant differences were demonstrated by chi square across patient-level categories comparing goiter subtypes (p<0.05). Notably, Whites had their highest frequency among thyroiditis, Blacks had their highest frequency among SG.  Patients aged > 65 years had their highest frequency among SG while <45 had their highest frequency among thyrotoxicosis. Women predominated overall, but men had their highest frequency among SG closely followed by cancer. Medicare patients’ highest frequency was among SG, which was the lowest frequency diagnosis for privately insured patients.  Patients in the <25th income percentile had their highest frequency among SG, which was the lowest for 51-75th and 76-100th percentiles.  Hospitals in the South had their highest frequency among SG, while those in the Northeast had their highest frequency among cancer. Multivariable adjusted logistic regression models found the factors associated with operation for SG were: age > 65 (AOR 1.82 [CI 1.45-2.27]) and 45-65 (AOR 1.34 [1.15-1.55]), black race (AOR 1.38 [1.18-1.62]), self-pay (AOR 1.96 [1.42-2.69]), hospital location in the South (AOR 1.5 [1.15-1.96]) and Midwest (AOR 1.53 [CI 1.15-2.05]). Odds were decreased for women (AOR 0.69 [0.6-0.8]); incomes 26th-50th % (AOR 0.82 [0.7-0.97]), 51st-75th % (AOR 0.67 [0.56-0.82]), and 76th-100th % (AOR 0.64 [90.5-0.82]).

Conclusion:
This study elucidates distributions within goiter types undergoing thyroidectomy in the U.S. SG associated with iodine-deficiency remains a national issue that is disproportionately found in patients who are older, poorer, of black race, or are treated in the South and Midwest. Identifying these risk factors may allow for targeted efforts to reduce SG nationwide.

43.10 Survival Disparities In Patients With Pancreatic Neuroendocrine Tumors

Posted on January 18, 2018

J. Ousley1, J. A. Castellanos1, C. E. Bailey1, N. Baregamian1  1Vanderbilt University Medical Center,General Surgery,Nashville, TN, USA

Introduction: The incidence of pancreatic neuroendocrine tumors (PNET) is rising, but little is known about the impact of disparities on survival. The aim of our study was to characterize existing disparities in patients with PNET.

Methods:  A retrospective cohort study of patients diagnosed with PNET was performed using the Surveillance, Epidemiology, and End Results (SEER) database between 1988-2012.  Kaplan-Meier and log-rank test were used for survival analysis.  A multivariable (MV) logistic regression model was used to assess demographic and tumor-related factors associated with survival.  

Results: A total of 3,759 patients with PNET were identified.  The mean age at diagnosis was 57.7 + 13.9 years.  The majority of patients were male (54%), white (80.5%), married (64.2%), and presented with distant disease (55.4%).  Median overall survival (OS) for the entire cohort was 52 months.  Median OS was significantly improved for patients who underwent primary tumor resection (PTR) compared to those who did not (64 vs 29 months, p<0.001).  On MV analysis, increasing age, male sex, higher grade tumors, and increasing stage were associated with worse survival, whereas tumor location in the tail of the pancreas and PTR were associated with improved survival (Table).  

Conclusion: Significant survival disparities were found in a cohort of patients with PNET, with improved survival observed in patients who underwent primary tumor resection.  Future studies focusing on access to care, patient education, and socioeconomic factors may help elucidate key factors for improved survival in patients with PNET.
 

43.08 Lymph Node Harvest and Survival in Gastric Cancer: A Single Institution Experience

Posted on January 18, 2018

C. Rog1, L. Brubaker1, C. Hsu1, E. Silberfein1  1Baylor College Of Medicine,Surgical Oncology,Houston, TX, USA

Introduction:  Nodal stage remains one of the most important prognostic factors for patients with gastric cancer and the National Comprehensive Cancer Network (NCCN) recommends that a minimum of 15 lymph nodes be resected for accurate staging.  Furthermore, recent studies suggest that the number of nodes resected may correlate with survival.  We examined the oncologic outcomes of lymph node harvest in an inner city, indigent, minority population with gastric cancer.

Methods:  A retrospective consecutive cohort study of 71 patients undergoing gastrectomy with curative intent at a single county hospital between 2001 and 2017 was performed.  Patients were divided into three groups based on the number of lymph nodes harvested.  Groups 1-3 included fewer than 15 nodes resected, between 15 and 25 nodes resected, and greater than 25 nodes resected, respectively.  Demographic, pathologic, and survival data were compared between groups.

Results: There were 12, 17 and 32 patients in groups 1-3 respectively.  The median numbers of nodes resected in each group were 12, 16 and 31 respectively.  There were no differences in median age, sex, type of surgery (total vs subtotal gastrectomy), tumor grade, pathologic T stage, node positive rate or clinical stage between groups.  Patients in group 3 had significantly higher N stage (p=0.03).  At a median follow up of 1036 days, there were no significant differences in survival for all stages between the 3 groups (p=0.125). 

Conclusion: We did not observe significant differences in survival with respect to number of lymph nodes resected for gastric cancer patients at our institution.  This may not be surprising, however, as the surgical procedures performed for these patients remained fairly consistent, and therefore the cohorts were stratified by the number of nodes identified by the pathologist as opposed to any inherent difference in therapy.  Likewise, our data do not suggest that more precise staging improved survival outcomes.  Retrospective studies correlating lymph node count with improved survival should be interpreted with caution.

 

43.07 Thyroid Cancer Disparities Among Native and Foreign-Born Hispanics in South Florida

Posted on January 18, 2018

O. Picado1, A. R. Marcadis1, E. N. Kobetz2, R. R. Balise3, J. I. Lew1  1University Of Miami Leonard M. Miller School Of Medicine,Division Of Endocrine Surgery, DeWitt Daughtry Family Department Of Surgery,Miami, FL, USA 2University Of Miami Leonard M. Miller School Of Medicine,Department Of Medicine,Miami, FL, USA 3University Of Miami Leonard M. Miller School Of Medicine,Division Of Biostatistics, Department Of Public Health Sciences,Miami, FL, USA

Introduction:  Recent trends suggest increasing incidence of thyroid cancer among all racial/ethnic groups in the United States. Hispanics are the largest racial/ethnic group in the country, representing 17% of the population. South Florida has an exceptionally high concentration (41%) of Hispanics composed primarily of Cubans, South and Central Americans. This study examines the differences in thyroid cancer rates and disease burden of Hispanics in South Florida.

Methods:  A retrospective review of 1012 patients who underwent thyroidectomy from 2010 to 2016 at a single academic institution was performed. Demographics including age, sex, race/ethnicity, and country of origin were reviewed. Racial/ethnic groups were defined as white (non-Hispanic), black (non-Hispanic) and Hispanic. All patients underwent pre-operative thyroid ultrasound and fine needle aspiration (FNA). Pathologic characteristics of thyroid cancer including histology, tumor size, multifocality, extrathyroidal and lymphovascular invasion were reviewed. Descriptive statistics and subgroup analysis were performed among Hispanic patients in relation to country of origin.

Results: Of 529 patients with thyroid cancer, 63% were Hispanic (n=335), 28% white (n=147) and 9% black (n=47). Hispanic patients were more likely to be foreign-born compared to white and black patients (61% vs 28% vs 31%, p<0.01). Hispanics with thyroid cancer were younger than their white and black counterparts (respectively, 45 vs 47 vs 51 years, p<0.001), had smaller nodules (2.1 vs 2.1 vs 2.8 cm, p<0.01), and more worrisome ultrasound features including irregular borders (41% vs 36% vs 19%, p<0.05) and microcalcifications (45% vs 46% vs 19%, p<0.05). FNA results were more commonly malignant in Hispanic and white patients compared to black patients where FNA results were mostly benign (p<0.01). While papillary thyroid carcinoma (PTC) was the most common thyroid cancer in all groups, foreign-born Hispanics had more diffuse (7% vs 4%) and tall cell (9% vs 3%) variants of PTC compared to native born Hispanics. There was also a higher proportion of follicular thyroid cancer (5% vs 2%) among foreign-born Hispanics. Stratified analysis of foreign-born Hispanics by region showed lower rates of PTC among Cubans (88%) and South Americans (86%) with higher rates of follicular carcinoma and medullary thyroid cancer, respectively.

Conclusion: South Florida encompasses a diverse patient population as demonstrated by a varied presentation of thyroid cancer among different racial/ethnic groups including Hispanics. Differing thyroid cancer incidence, clinical presentation and tumor histopathology among Hispanic patients may be associated with environmental or behavioral factors, especially among those that are foreign-born. Clinicians should be prepared and culturally competent to provide care to this fastest growing racial/ethnic group in the United States.  

 

43.04 African American race associated with higher cost of surgical care for thyroidectomy

Posted on January 18, 2018

S. Jang1,2, C. J. Balentine1, H. Chen1  1University Of Alabama at Birmingham,Surgery,Birmingham, Alabama, USA 2Howard Hughes Medical Institute,Chevy Chase, MD, USA

Introduction: Racial disparities in health care and health outcomes have been well documented in most diseases, but there is limited data for thyroid disease. Thyroidectomy is the mainstay for many thyroid diseases, but its cost among different racial and ethnic groups are largely unexamined. The purpose of this study was to examine the association between race and ethnicity and the total hospital cost of thyroidectomy.

Methods: This retrospective study included 898 consecutive cases in our institution between September 2011 and July 2016 coded as complete thyroidectomy or total thyroidectomy using ICD-9 and 10 procedure codes. We evaluated demographics, insurance type, and readmission rates. Total length of stay and hospital costs were evaluated using the Mann-Whitney U and the Kruskal-Wallis non-parametric tests. Categorical variables were evaluated with chi-square. Distributions are denoted by standard error of the mean.

Results: The study population was 64.0% Caucasian, 33.2% African American, 0.8% Hispanic, and 0.5% Asian. Median age was 48 years, 81% were female, and 77% were outpatients. Blue Cross was the most common payer type (50%). Total hospital costs were greater for African American patients ($6,750.23 ± 372.15, p<0.001) compared to Caucasian patients ($5,890.87 ± 233.96) but not for Hispanic and Asian patients. Compared to Caucasian patients, African American patients were more likely to experience hospital costs greater than $10,000 (9.1% vs 4.7%, P = 0.007), and the difference in total cost was still significant even after the exclusion of cases that cost above $10,000. Mean length of stay was 1.61d ± 0.20  for African American patients while it was 0.93d ± 0.07 for Caucasian patients (P <0.001), where 71% and 54% were discharged on the same day, respectively (p<0.001). Nevertheless, there were no difference in readmission rate between African American and Caucasian patients (P = 0.958). Additionally, African American and Caucasian males had higher costs (p<0.001 and trending at 0.054, respectively) and longer length of stay (p<0.001 and 0.047, respectively) compared to their racial counterparts. Comparing specific costs across the four groups showed that African American males had the highest cost of anesthesiology (P=0.001) and Caucasian females had the lowest cost of labs (p<0.001). There were no difference in cost of radiology and heart center across the groups (p=0.078 and 0.558, respectively).

Conclusion: African American race was associated with higher hospital costs for thyroidectomy compared to Caucasian patients. The increased cost could be explained in part by longer length of stays after the operation. Thyroidectomies done on male patients were more costly compared to their female racial counterparts. Examining specific areas of racial disparity in surgical cost is a potent method of addressing economic and social inequality and can potentially reduce cost of health care.

43.06 The Urban Versus Rural Divide for Completion Thyroidectomy: A Profile From Two States

Posted on January 18, 2018

J. R. Imbus1, Y. Shan1, N. Brys1, G. Leverson1, J. Havlena1, N. Zaborek1, S. Pitt1, D. F. Schneider1  1University Of Wisconsin,Surgery,Madison, WI, USA

Introduction:

Following thyroid lobectomy, patients may require a completion thyroidectomy (CT) to remove the remaining thyroid tissue.  To avoid the possibility of CT, patients may prefer or be advised to undergo total thyroidectomy (TT) instead. The frequency of CT and associated risks are not well established, but potentially influenced by patient- and hospital-level factors. The purpose of this study is to determine the incidence, factors, and complications associated with CT.

 

Methods:

We identified adult patients who underwent thyroidectomy for benign disease using the Healthcare Cost and Utilization Project State Inpatient Databases and State Ambulatory Surgery and Services Databases for Florida and New York from 2011-2013. We defined the CT cohort as patients undergoing two thyroid operations within 133 days (85th percentile). Multilevel logistic regression identified hospital and patient-level factors associated with undergoing CT (vs definitive lobectomy or TT), as well as factors associated with complications occurring within 30 days post-operatively.

 

Results:

Of all 37,282 patients (definitive lobectomy, CT, and TT), 4.6% (n = 1,713) underwent CT. Compared to urban hospitals, CT frequency was higher in rural hospitals (7.3% vs 4.5%, p < 0.001) whereas TT was less frequent (21.4% vs 39.9%, p<0.001). CTs were less common in black patients (3.1% vs 4.7%, p <0.001) compared to white patients. Black patients underwent more TTs (47% vs 37%, p < 0.01) compared to white patients. After adjustment for patient and hospital-level factors, rural hospitals (vs urban hospitals) were associated with CT (OR 1.77, 95% CI 1.16-2.71, p < 0.01). Compared to whites, blacks were less likely to undergo CT (OR 0.66, 95% CI 0.56-0.79, p < 0.001).

 ¶ Rates of hematoma (0.8% vs 1.3%, p=0.1) and RLN injury (0.8% vs 0.8%, p=0.99) in CT did not differ from TT, but hypocalcemia was less frequent (5.6% vs 7.9%, p<.001). Additional multilevel modeling revealed that post-operative complications (hypocalcemia, RLN injury, or hematoma) were associated with Charlerson co-morbidity index greater than 1 (p<0.001) for patients undergoing CT or TT. In this cohort, complications were nearly 70% less likely for outpatient operations (OR 0.28, 95% CI 0.24-0.33, p<0.001,). Age over 45 years was associated with decreased odds of complications (p<0.001), and patients undergoing CT were less likely to have a complication compared to patients undergoing TT (OR 0.79, 95% CI 0.63-0.98). Black patients undergoing either CT or TT were twice as likely to suffer a RLN injury compared to white patients (OR 2.03, 95% CI 1.32-3.13, p<0.01).

 

Conclusion:

In two large states, the rates of CT are expected based on thyroid cancer incidence and treatment guidelines. Higher rates of CT in rural hospitals appears related to less extensive use of initial TT. More frequent index TT may explain lower rates of CT in black patients, as well as their vulnerability to RLN injury. 

43.01 Decision Making Roles among Diverse Patients with Cancer

Posted on January 18, 2018

A. De Roo1, M. Miller2, A. Antunez1, C. Kin2, A. Morris2  1University Of Michigan,General Surgery,Ann Arbor, MI, USA 2Stanford University,General Surgery,Palo Alto, CA, USA

Introduction:  Collaborative decision making (DM) between patients and physicians is an essential feature of patient centered care. Patient DM roles are categorized as (1) active or patient-driven; (2) collaborative, in which the patient and physician together form a treatment plan; and (3) passive or physician-driven. We hypothesized that lower socioeconomic status and minority ethnicity are associated with a more passive role in cancer care DM. 

Methods:  To investigate preferred and actual DM roles for cancer care, we performed a systematic review of the literature in 5 databases. The search string encompassed “decision making”, “cancer care” and “socioeconomic status” and was limited to English language articles published between 2006-16. One reviewer assessed 1454 abstracts and excluded 1375 that did not pertain to cancer treatment, adult patients, primary data, or peer review. The remaining 79 articles were reviewed in full by >= 2 study team members to verify final eligibility: comparison of preferred or actual DM role among two or more socioeconomic groups or ethnicities. We developed a data abstraction tool to capture study features, population, methods, results, conclusions, and methodological quality. Data from quantitative and qualitative articles were integrated through constant comparison across a joint display table and triangulated through iterative discussion among the team.

Results: 24 quantitative, 2 mixed methods, and 6 qualitative studies met inclusion criteria. Among most studies, older age, low health literacy and low income were associated with a more passive role in DM. Low-acculturated Latina women were more likely to report poor understanding of treatment options and, consequently, to prefer a more passive DM role. They were also less likely to be satisfied with DM role and with care when they did not understand the rationale for the treatment decision.  In contrast, high-acculturated American Latina women were more active decision makers, as were younger patients and those with breast cancer. Although lower satisfaction scores were noted in patients from minority populations or with low health literacy, American Indians and African Americans indicated that trust and being treated with respect were associated with DM role satisfaction and adherence to therapy.

Conclusion: Patients with lower health literacy, socioeconomic status or non-majority demographics were more likely to play a passive role in cancer treatment decision making, which was associated with lower satisfaction especially if they did not understand the rationale for treatment decisions. Improved patient education materials and interactions for a variety of languages and health literacy levels will allow these patients participate as desired in their cancer treatment choices. With the information gained from this study, care provider training and patient education materials can be adjusted to allow for greater patient collaboration in decision making.

43.02 Why is Hyperparathyroidism Under-diagnosed and Under-treated in Older Adults?

Posted on January 18, 2018

A. Dombrowsky1, B. Borg1, R. Xie1,2, J. K. Kirklin1,2, H. Chen1, C. Balentine1,3,4  1University Of Alabama at Birmingham,Department Of Surgery,Birmingham, Alabama, USA 2Kirklin Institute For Research In Surgical Outcomes,Birmingham, AL, AL, USA 3University Of Alabama at Birmingham,Institute For Cancer Outcomes & Survivorship,Birmingham, Alabama, USA 4Birmingham & Tuscaloosa Veteran’s Affairs Hospital,Birmingham, AL, USA

Introduction:

Hyperparathyroidism significantly decreases quality of life, yet elderly patients are under-diagnosed and under-treated even though parathyroidectomy offers definitive cure with minimal morbidity. The purpose of this study is to determine why patients aged 75 years and older with hyperparathyroidism fail to be appropriately diagnosed and referred for surgical treatment.

Methods:

We reviewed charts for a random sample of 50 patients aged 75 and older who had hyperparathyroidism (elevated calcium and parathyroid hormone) and were referred (N=25) or not referred (N=25) for surgical evaluation. Two independent reviewers examined clinical notes and diagnostic data from medical records to identify reasons for delay in diagnosis of hyperparathyroidism and reasons for referral (or non-referral) for parathyroidectomy.

Results:

The mean age of our cohort was 84 ± 4 years, 90% were women, and 60% were Caucasian. Mean follow up was 6.2 ± 4.8 years.

        Reasons For Delay Or Missed Diagnosis: In 58% of patients, an elevated serum calcium was not acknowledged over the course of 257 distinct physician encounters. When hypercalcemia was noted, it was frequently attributed to other causes including calcium supplements (18% of patients), diuretic use (12%), dehydration (10%), renal dysfunction (10%), and vitamin D deficiency (16%). Even when calcium and parathyroid hormone levels were both elevated, the diagnosis was missed in 28% of patients, and 16% with clear symptoms of hyperparathyroidism remained undiagnosed.

        Reasons for Lack of Surgical Referral: In 42% of patients, a non-surgeon decided at some point in their care plan that surgery offered no benefit. Additionally, the decision to reject surgery as a treatment option was made for 36% of patients despite their developing new symptoms of hyperparathyroidism (fracture, osteoporosis, kidney stones) or having an increase in serum calcium by > 1 mg/dl.  Of the 25 patients who were referred to surgeons, 1 patient decided not to follow up, and 4 patients stated they did not want surgery.  Of the 20 patients who saw a surgeon, 17 (85%) were deemed eligible for surgery by their surgeon.

Conclusion:

Substantial gaps exist in processes for diagnosis and referral of patients with hyperparathyroidism that lead to under-diagnosis and under-treatment.  To improve rates of diagnosis and treatment of hyperparathyroidism, strategies are needed to educate non-surgeons and patients about the benefits of surgery and to modify care processes to more efficiently diagnose and refer patients.   Patient quality of life can be substantially improved by reducing the time to definitive treatment of hyperparathyroidism.

43.03 Pelvic Exenteration for Locally Advanced Rectal Cancer in a Safety Net Hospital

Posted on January 18, 2018

A. Zimmerman1, E. Rachlin1, C. Rog1, C. Hsu1, E. Silberfein1  1Baylor College Of Medicine,Surgical Oncology,Houston, TX, USA

Introduction: Rectal cancer with local invasion of adjacent pelvic organs is a difficult problem requiring multidisciplinary care. Pelvic exenteration, or the en-bloc resection of the neoplasm and associated viscera, is a procedure that offers the possibility of curative resection often at the expense of considerable morbidity. We are fortunate to be able to offer this procedure in our resource-limited, county hospital setting. Outcomes of these operations in underserved populations have not been explored.

Methods: A retrospective consecutive cohort study of 17 patients undergoing total or posterior pelvic exenteration for locally advanced rectal cancer at a single county hospital between 2010 and 2017 was performed. Demographic, pathologic, recurrence, and survival data were analyzed utilizing summary statistics.  Collaboration among different surgical teams was assessed.

Results: The median age of the cohort was 54 years.  The majority (71%) were women and of Hispanic ethnicity (41%).  All but one of the patients underwent neoadjuvant treatment.  Twelve patients (71%) underwent posterior pelvic exenteration and 5 underwent total pelvic exenteration.  Four different teams of surgeons were involved in these operations including members from surgical oncology, gynecologic oncology, plastic surgery, and urology. Eight of the cases involved three or more teams of surgeons, and nine cases were performed by two teams.  An R0 resection was accomplished in 13 patients (76%).  The remaining patients underwent an R1 resection.  The median length of stay was 11 days.  The 30-day morbidity was 24% and overall there were 9 complications in 8 patients.  All but 2 complications were Clavien-Dindo class I or II.  At a median follow up of 986 days there were 3 local recurrences and 4 distant recurrences.  There were 2 patient deaths at 875 and 1190 days post-op, respectively.

Conclusion: The outcomes for the treatment of locally advanced rectal adenocarcinoma at a safety net hospital with limited resources are favorable compared to contemporary reports in the medical literature. R0 resection remains the best indicator for mortality following pelvic exenteration. The 76% R0 rate at our institution is well above the published average, which ranges from 37-57%.  This may reflect a high proportion of patients in our population with localized but neglected malignancies in contrast to aggressive biology that may be seen in insured patient populations.  Despite the relative lack of resources at a county hospital, collaboration among surgical teams can be attained and pelvic exenteration can be performed safely with favorable outcomes.

 

42.18 MicroRNAs Present In CSF As Biomarkers Of Spinal Cord Ischemia And Paresis After TEVAR

Posted on January 18, 2018

H. F. El Sayed1, E. Tili2, J. Michaille3,4, H. E. Awad2  1Ohio State University,Department Of Surgery,Columbus, OH, USA 2Ohio State University,Department Of Anesthesiology,Columbus, OH, USA 3University Of Bourgogne,Dijon, , France 4Ohio State University,Department of Cancer Biology And Genetics,Columbus, OH, USA

Introduction:
Approximately 24,000 new cases of thoracic aortic aneurysms are diagnosed each year in the USA. Spinal cord ischemia (SCI) and paralysis is a devastating complication following thoraco abdominal aortic aneurysm (TAAA) open and endovascular repair. The rates of paralysis after TAAA repair range 5-15%, even with the adjunctive non-pharmacological preventive strategies such as cerebrospinal fluid (CSF) drainage and blood pressure management. However, our current strategies are not sufficiently effective to prevent and/or treat SCI after TAAA repair. Furthermore, the rates of paralysis in new interventions, such as thoracic endovascular aortic repair (TEVAR), the risk of paralysis is not significantly lower than open repair methods. Currently, there is no standard of- care pharmacological preventive treatment and, to our knowledge, no molecule is being tested for this purpose. Paraplegia is one of the most feared and devastating complications following TAAA and endovascular repair. Patients who developed paralysis after TEVAR suffer from high morbidity and mortality. 

Methods:

We are working to identify microRNA signatures present in the CSF samples of patients undergoing TAAA repair, with and without paresis/paralysis. We aim to identify microRNAs that are associated with paresis/paralysis and that can be used as biomarkers of this event.  Our study is limited to those patients where a spinal drain for CSF drainage is inserted (not emergency patients) to allow us repeated CSF sampling. Samples are being  collected before surgery and 6, 12, 24, and 48 hours after surgery in all patients. Additional samples are collected in cases with neurological complications (ie. Patients that develop delayed paresis and come back at the hospital, as it was the case with the last patient) to differentiate between microRNAs present in the CSF of paralyzed vs non-paralyzed patients, as well as before and after TEVAR. We are also planning to run similar experiments with the plasma of these patients, to see if we will be able to identify a blood microRNA signature related to spinal cord ischemia.  This method is less invasive and can be used for all patients, regardless of the spinal  drain availability.

Results:

CSF samples are the analyzed for microRNA expression using NanoString technology arrays. Candidate microRNAs are further validated by qRT-PCR techniques using Taqman probes. Our current preliminary data make us confident that we will be able to identify microRNAs that can be used as biomarkers of paresis/weakness event. 

Conclusion:

Patients that show the presence of the identified candidate microRNAs in their CSF, will be considered at high risk or “predisposed” to develop paresis and as such these patients should be followed up after they leave the hospital. We posit that the identified microRNAs can be further used for development of future therapeutics aimed at prevention of paralysis.

42.19 Nasal Delivery of VEGF Increases Shedding of HB-EGF in a Murine Model of Compensatory Lung Growth

Posted on January 18, 2018

J. T. Vuong1,2, D. T. Dao1,2, A. Pan1,2, L. Anez-Bustillos1,2, D. Bielenberg1,2, M. Puder1,2  1Boston Children’s Hospital,Department Of Surgery,Boston, MA, USA 2Boston Children’s Hospital,Vascular Biology Program,Boston, MA, USA

Introduction:  

Vascular endothelial growth factor (VEGF) plays a key role in angiogenesis with implications for organ development and regeneration. Our group has previously demonstrated that systemic VEGF administration accelerated compensatory lung growth (CLG) in a murine model of left pneumonectomy. The aims of this study are to determine the mechanism of how VEGF accelerates CLG and to explore nasal delivery as an alternative route for VEGF administration.

Methods:

Eight-week old C57Bl6 male mice underwent left pneumonectomy, followed by daily nasal instillation of VEGF at 0.5 mg/kg (N=10) or isovolumetric saline (N=10). Mice were euthanized on postoperative day (POD) 4, and lungs were harvested for immunohistochemistry and protein expression analysis.

To determine the mechanisms of VEGF-mediated acceleration of lung growth, human lung microvascular endothelial cells (HMVEC-L) were treated with or without VEGF (10 ng/mL) for 24 hours. Conditioned media were then placed on human bronchial epithelial cells (HBEC). Viability of HBEC cells was assessed every 24 hours for 72 hours through a colorimetric assay. Protein contents of HMVEC-L conditioned medium were characterized with an angiogenesis array.

Results:

Compared to saline-treated mice, inhaled VEGF treatment resulted in increased expression of heparin binding EGF-like growth factor (HB-EGF) (P = 0.02) and activation of its main receptor, epidermal growth factor receptor (P = 0.03), which promotes epithelial cell proliferation and differentiation (Figure 1A). However, there was no difference in proHB-EGF expression on immunohistochemistry, indicating that the increase in HB-EGF levels was due to increased shedding of the factor. Mice in the VEGF group displayed a lower Ki-67 proliferation index (P=0.03) on POD 4, reflecting that cellular proliferation was near completion in VEGF-treated mice and that VEGF accelerated compensatory lung growth.

Conditioned medium from VEGF-treated HMVEC-L increased HBEC viability (P=0.046) and contained a 6.5-fold increase in HB-EGF level compared to VEGF-negative conditioned medium (Figure 1B).

Conclusion:

Nasal delivery of VEGF may accelerate compensatory lung growth via increased shedding of heparin binding EGF-like growth factor. 

 

 

 

42.16 A Computational Study of Papillary Muscle 3D Kinematics to Evaluate Mitral Valve Motion

Posted on January 18, 2018

J. Park2, P. Bonde1  1Yale University School Of Medicine,Cardiac Surgery,New Haven, CT, USA 2Yale University School Of Medicine,Bonde Artificial Heart Lab,New Haven, CT, USA

Introduction:
Current understanding of myxomatous mitral valve prolapse is limited to longitudinal echocardiographic studies with indicators of left ventricular adverse remodeling. This has resulted in surgical interventions being instituted when left ventricular dilatation has reached a certain dimension. Fundamental understanding of the mitral valve mechanism may provide a therapeutic target to be instituted earlier. In this study, we performed a computational study on the kinematics of the two-papillary muscles manipulating mitral valve motion.

Methods:
Three dimensional model of mitral valve is considered as a body at one end with two massless rods connected to the other end representing annular leaflet and chordal attachments. In order to evaluate spatial motion of each papillary muscle in three dimensions and determine its physiologically working range in cardiac cycle, kinematic analysis was carried out. Two massless rods within six degrees of freedom were simulated by varying its thickness and orientations based on spherical coordinate system (r, θ, φ), where r is thickness of the papillary muscle, θ is polar angle, and φ is azimuthal angle. Assuming that papillary muscle thickness variations can represent annulo-papillary muscle distance, three different thicknesses of 10, 15, and 25 mm were used in the simulation. Based on the plausible interpapillary muscle distance variations defined as 15 to 30 mm, three dimensional working ranges of the two papillary muscles for each condition were visualized.

Results:
Possible combinations were evaluated with a step interval of π/10. Assuming that two papillary muscles approaching or receding to each other within 45 degrees at maximum from their vertical position in systole and diastole, the distance between the fixed points of them are set to be 22.5 mm. For the simulation case with the papillary muscle thickness of 10 mm, 4961 out of 10201 cases, (48.6%) were found to be effective orientations for creating plausible interpapillary muscle distance from 15 to 30 mm. With larger thickness for both (r1, r2 = 25 mm), effective cases were reduced to 30% (2961 out of 10201 cases) and the working range overlaps up to the fixing position of each other, meaning fixing distance of the papillary muscle needs to be properly set based on their spatial movement. Within different thicknesses, 10 mm for the one papillary muscle (r1 = 10 mm  and 25 mm for the other papillary muscle (r1 = 25 mm , asymmetric working zone could be seen. 

Conclusion:
Three dimensional kinematic analysis on two papillary muscles offers unique insights in to the working ranges and interactions based on their geometries and configurations. This model offers a foundation for an advanced dynamic model of mitral valve.
 

42.17 Assessment of Differential Pressure Levels in Chest Drainage Systems

Posted on January 18, 2018

V. V. Dobaria1, R. Cameron1, E. Aguayo1, K. Bailey1, Y. Seo1, Y. Sanaiha1, P. Benharash1  1University Of California – Los Angeles,Cardiac Surgery,Los Angeles, CA, USA

Introduction:

Chest drainage systems are routinely used to achieve proper drainage of the pleural space after cardiac and thoracic operations. The ideal chest drainage system would evacuate air and fluid, creating a negative intrapleural pressure that would facilitate lung expansion and induction of adhesions to prevent accumulation of matter in this potential space. While the appropriate level of suction that would create the most rapid resolution of air leaks and pleural collections has been debated, data on the actual intrapleural pressure during the use of chest drainage systems is lacking. Columns of fluid collecting in the drainage tube may alter the transmission of optimal pressure to the patient. The present study was performed to evaluate differences along the length of the chest drain circuit using an ex-vivo model.

Methods:  

An ex-vivo apparatus coupled to a commercial pleural drainage system was devised to provide calibrated manometer measurements and user-defined levels of suction and air leak. Simultaneous pressure measurements were obtained at the outlet of the drainage system and at the simulated entry site into the patient where air could be introduced. Systematic trials were conducted with increasing levels of water between the patient and drainage modules and at various levels of suction and air leak rates.  Signals were recorded at 100Hz on a PC and analyzed using two-way t-tests. Significance was taken at P-value < 0.05.

Results:
With no obstruction, the pleural drainage system provided precise levels of negative pressure at the patient level (10-40 cm H2O). Addition of fluid in the drainage tubing as a water trap caused significant differences in levels of transmitted suction as shown in Figure. With increasing air leakage and fluid volume, the pressure differential between the drainage system and patient increased significantly (1.14 ± 0.04 to 36.69 ± 0.15cm H2O, P<0.001). While off suction increasing levels of obstruction from 0 to 22 cm of water, led to a development of positive intrapleural pressures varying from 2.6 ± 0.4 to 12.1 cm H2O, P< 0.001).

Conclusion:

The findings of the study suggest that while commercially available pleural drainage systems provide predictable levels of suction at the level of the device, intrapleural pressures can be highly variable and depend on complete patency of connecting tubes. In fact, high levels of positive pressure may develop and precipitate lung collapse and therefore prolong the duration of air leaks. This finding may explain the highly variable rates of reported postoperative air leak. Systems able to modulate the level of suction based on actual intrapleural pressures may enhance recovery after procedures requiring tube thoracotomy. 

42.14 An Endovascular, Totally Implantable Cardiac Support Device for early Class III Heart Failure

Posted on January 18, 2018

J. Park1, P. Bonde2  1Yale University School Of Medicine,Bonde Artificial Heart Lab,New Haven, CT, USA 2Yale University School Of Medicine,Cardiac Surgery,New Haven, CT, USA

Introduction:
Early class III heart failure patients can benefit greatly from increase in their cardiac output, provided they do not suffer from the adverse events related to current left ventricular assist device (LVAD)s, such as stroke, drive line infections and pump thrombosis.  We present an endovascular, totally implantable, on demand cardiac support device with technological sophistication to eliminate common adverse events associated with LVADs.

Methods:
Wirelessly controllable on-demand recover device is composed of four elements: axial pump, two valves, cleaning port, and transcutaneous energy transfer system (TETS). Brushed DC motor (8.5 mm in diameter and 20 mm in length) capable of operating under the maximum unload speed of 52kRPM at 7.4 V was used for manufacturing axial pump in which impeller and diffuser designs were qualified by computational fluid dynamics (CFD). Gear-driven valve was designed to generate open and close maneuver with proper amount of force applied. Objet 30 3D printer with resin material was used to produce watertight prototypes. Performance tests were carried out using mock circulation loop equipped with flowmeter and pressure transducers. Color dyes were used to quantitatively evaluate valve sealing capability and cleaning efficiency. Pump performance and wireless power transfer efficiency were also evaluated by creating pump performance curve that shows flow and pressure relationship and measuring the power successfully delivered to the system respectively.

Results:
Pump performance curve was created by controlling flow with manufactured valves. With two valves closed at the end of the graft where pump are located inside, colored dyes were introduced into the graft and leaking dye concentration was measured, which results in negligible value (< 0.1 ppm). Dyes inside the graft were then cleared out through cleaning ports. Concentration of the dye inside the graft after cleaning successfully dropped near zero (< 0.1 ppm). Pump and valve operations were all achieved by wireless power delivery with an efficiency of 80 %. The device is planned to pump blood from foramina ovale and return to ascending aorta, with ability to shut the inflow and outflow and clean the pump automatically when desired (7 cc volume), thus eliminating any clot formation within the pump.

Conclusion:
An endovascularly deployable, on-demand cardiac support device was successfully tested in-vitro. With its ability to periodically clean and maintain the pump, it eliminates clot formation. Total implantability has the potential to offer better quality of life and acceptability for the early stage class III heart failure patients.
 

42.15 Impaired Coronary Contractile Response to Phenylephrine after CP/CPB in Diabetic Patients

Posted on January 18, 2018

C. Gordon1, J. Feng1, N. Sellke1, I. Lawandy1, A. Gorvitovskaia1, Y. Liu1, A. Ehsan1, J. G. Fingleton1, F. W. Sellke1  1Rhode Island Hospital,Cardiothoracic/Surgery,Providence, RI, USA

Introduction: We have previously found that cardioplegic arrest and cardiopulmonary bypass (CP/CPB) is associated with impairment of coronary arteriolar response to phenylephrine in non-diabetic (ND) patients.  In this study, we further investigated the effects of diabetes on coronary arteriolar response to phenylephrine in the setting of CP/CPB and cardiac surgery.

Methods: Coronary arterioles (90-150 µm in diameter) were harvested pre- and post-CP/CPB from the ND and diabetic (DM) patients (n = 8/group) undergoing cardiac surgery. In-vitro microvascular reactivity was examined in the presence of phenylephrine. The protein expression/localization of the alpha-1 (α -1) adrenergic receptors in the atrial myocardium were measured by Western Blotting and immunohistochemistry.

Results: Phenylephrine (10-9-10 -4 M) induced dose dependent contractile responses in both ND and DM vessels pre and post-CP/CPB. There were no significant differences in the contractile responses to phenylephrine between pre-CP/CPB ND and DM vessels. The post-CP/CPB contractile responses were significantly diminished compared to the pre-CP/CPB in the two groups (P<0.05 vs. pre-CP/CPB). This diminished contractile response was more pronounced in the post-DM vessels than in the post-ND vessels (P<0.05 vs. ND). There were no significant differences in the protein expression of  α-1 receptors in the atrial myocardium between the ND and DM tissue or between pre-CP/CPB and post-CP/CPB tissue.

Conclusion: Diabetes is associated with a decreased contractile response of coronary arterioles to phenylephrine in the setting of CP/CPB.  This alteration may contribute to the vasomotor dysfunction of coronary microcirculation seen early after CP/CPB in diabetic patients.
 

42.12 Association of Aortic Wall Hypoxia in Rhesus Macaques Exposed to Environmental Tobacco Smoke

Posted on January 18, 2018

E. S. Lee1,3, A. T. Nguyen3, K. J. Dolan3, Z. Irwin3, A. Rona3, K. M. Samadzadeh3, H. Smothers3, Z. Smit-McBride2, E. S. Lee1,3  1UC Davis,Surgery,Sacramento, CA, USA 2UC Davis,Eye Center,Sacramento, CA, USA 3Sacramento VA Medical Center,Surgery,Mather, CA, USA

Introduction:

Tobacco use is known to contribute to atherosclerosis and vascular disease. An intimate link exists between oxygen tension and inflammation within localized vascular tissue beds and tobacco smoke. However, we believe that vascular tissue response to environmental tobacco smoke is differential and that the infra-renal aorta is more prone to vascular tissue hypoxia. In this study, we sought to establish the presence of hypoxia in vascular segments of rhesus macaque monkeys after tobacco exposure

Methods:

Rhesus macaques were exposed to tobacco smoke exposure 6 hours per day, 5 days per week. Tobacco smoke exposure occurred over a 4 week period. Control monkeys were exposed to filtered air. Tissue and blood collection occurred on the day immediately following the last day of exposure. Cotinine levels were assessed in monkey's plasma via ELISA techniques. The infra-renal aorta and thoracic aorta were collected and vascular segments were snap frozen in liquid nitrogen. Total protein for hypoxia inducible factor-1 alpha (HIF-1α) was assessed using Western Blotting techniques. Comparisons were made between the thoracic aorta and the infra-renal aorta as well as the between aortae exposed to cigarette smoke and filtered room air. A Student’s 2-tailed t-test was used to determine differences and a P<.05 was considered significant.

Results:

Four monkeys were exposed to tobacco smoke and 2 monkeys served as controls. Serum cotinine levels were higher in monkeys exposed to cigarette smoke compared to controls after 30 days of exposure: 54.9±5.0 ng/mL vs. 0.06±.001 ng/mL (P<.01). HIF-1? was found to be expressed greater in the abdominal aortic tissue from monkeys exposed to tobacco smoke 1.08±0.09 compared to the aortic tissue from controls 0.842±0.11 (P=.04). Similarly, HIF-1? was found to be expressed greater in the thoracic aortic tissue from monkeys exposed to tobacco smoke 1.11±0.16 compared to the thoracic aortic tissue from controls 0.818±0.11 (P=.12). Overall in all aortic samples, there was a trend for the abdominal aortic tissue to have greater levels of HIF-1? than the thoracic aorta (Figure).

Conclusion:

These preliminary results establish the presence of hypoxia in the thoracic and abdominal aorta with just minimal exposure of 4 weeks of cigarette smoke. Smoking and hypoxia in the abdominal and thoracic aorta of rhesus macaques provides a framework for future investigations into atherosclerotic plaque formation in this animal model.

42.13 Increased Coronary Contractile Response to Serotonin in Pig with Metabolic Syndrome

Posted on January 18, 2018

I. Lawandy1, Y. Liu1, V. Pavlov1, L. Scrimgeour1, F. W. Sellke1, J. Feng1  1Brown University School Of Medicine,Cardiothoracic/Surgery,Providence, RI, USA

Introduction: Metabolic syndrome (MetS) is associated with alterations in coronary vascular smooth muscle and endothelial function. The current study examined the contractile  response of  the isolated coronary arterioles to serotonin  in  pigs with or without metabolic syndrome (MetS) and investigated the signaling pathways reponsible for serotonin-induced vasomotor tone.  

Methods: The MetS pigs were developed by feeding with a hyper-caloric, fat/cholesterol diet and the control animals (Lean) fed with a regular diet for 12 weeks (Lean: n = 8; MetS : n = 6). The coronary arterioles (90-180 micrometers in diameter) were dissected from the harvested pig myocardial tissues and the in-vitro coronary arteriolar response to serotonin was measured in the presence of pharmacological inhibitors.  

Results: Serotonin (10-9-10-5M) induced dose-dependent contractions of coronary resistant arterioles from both control (Lean) and MetS pigs. This effect was more pronounced in the MetS vessels as compared with that of controls (Lean)  [53% of contraction (serotonin 10-6M) in MetS vessels vs. 26% of contraction in control, P<0.05]. Serotonin-induced contraction of the MetS vessels was significantly inhibited in the presence of the phospholipase A2 (PLA2) inhibitor quinacrine (10-6M, 53% vs. 24%),  the cyclooxygenase (COX) inhibitor indomethacin (10-5M, 53% vs. 27%), the thromboxane-prostanoid (TP) receptor antagonist SQ29548 (10-6M, 53% vs.31%, P<0.05, respectively). 

Conclusion: MetS is associated with increased contractile response of porcine coronary arterioles to serotonin via activation of PLA2, COX and subsequent TXA2, suggesting that alteration of vasomotor function may occur at early stage of MetS and obesity.

42.09 Differential Expression of FAS and NKLAM in Patients Undergoing Carotid Endarterectomy

Posted on January 18, 2018

N. Harroun1, G. S. De Silva1, J. Kornbluth3, L. A. Sanchez1, M. A. Zayed1,2  1Washington University In St. Louis,Division Of Vascular Surgery,Saint Louis, MO, USA 2Veterans Affairs St. Louis Health Care System,Division Of Vascular Surgery,Saint Louis, MO, USA 3Saint Louis University School Of Medicine,Department Of Pathology,Saint Louis, MO, USA

Introduction:
Atherosclerosis is prevalent in western society. Understanding the processes of atherogenesis remain an area of intense investigation. Fatty acid synthesis and tissue inflammation have been demonstrated to influence atherogenesis and atheroprogression.  Fatty Acid Synthase (FAS) is a central enzyme responsible for de novo synthesis of long chain saturated fatty acids. Natural Killer Lytic-Associated Molecule (NKLAM) is an important inflammatory mediator that can alter cellular viability. We hypothesize that FAS and NKLAM are important molecular mediators of disease progression and are therefore differentially expression in variably diseased segments of carotid arterial plaque

Methods:
Using an IRB-approved, prospectively maintained vascular tissue biobank, we analyzed eight carotid endarterectomy plaque samples from patients undergoing carotid endarterectomy (CEA). CEA plaque was partitioned into maximally (Max) and minimally (Min) diseased segments.  Protein lysates were obtained for both Max and Min diseased segments and Western blot analysis was used to evaluate FAS and NKLAM protein expression levels.  Values are reported as a mean  SEM and as ratios of protein expression to a loading control (Caveolin).  Wilcoxon paired-match rank test was performed to determine any significant differences, with p < 0.05 considered significant.

Results:
Max diseased carotid segments demonstrated higher FAS expression compared to Min diseased segments (0.230.07 vs 0.05  0.02, respectively; p =0.04). Interestingly, NKLAM expression was visibly lower in Max diseased segments, though this value was not statistically significant.

Conclusion:
We have preliminarily observed significant differential expression of FAS and NKLAM, which are important disease mediators in Max and Min diseased carotid plaque segments.  These findings suggest that different plaque regions may be affected by different biological processes that can influence disease progression.
 

42.10 Maintaining Liposome Stability for In Vivo Translation

Posted on January 18, 2018

O. H. Grandas1, R. K. Fisher1, R. Dieter1, S. S. Kirkpatrick1, J. D. Arnold1, R. M. Buckley1, M. M. McNally1, M. B. Freeman1, D. J. Mountain1  1University Of Tennessee Graduate School Of Medicine,Department Of Surgery,Knoxville, TN, USA

Introduction: RNA interference (RNAi), or the use of short interfering RNA (siRNA) to transiently attenuate cellular protein expression, shows promise as a gene therapy technology to alter vascular pathology. However, the development of molecular nanocarriers is needed to overcome the unfavorable physiochemical properties that limit intracellular siRNA delivery. We have previously established cell penetrating peptide-modified neutral PEGylated liposome (CPP-PLP) as viable molecular nanocarriers in vascular smooth muscle cells. We aim to translate this molecular drug delivery platform for in vivo application in vascular injury models of disease, using liposomes stable when stored for off-the-shelf administration.  Here we aim to determine the optimal storage parameters for CPP-PLP stability and siRNA capsulate retention.

Methods:  CPP-PLPs were formed as previously described, injected with 200nM siRNA, and stored at room temp, 4°C, or -20°C for 0-48hr in phosphate buffered saline (PBS) or Dulbecco’s Modified Eagle Medium (DMEM).  Additionally, 0.1% bovine serum albumin (BSA) in PBS was tested as a stabilizer protein and 0.1% dimethyl sulfoxide (DMSO) was tested in DMEM as a cryoprotectant. Dynamic light scattering (DLS) was used to measure nanoparticle size, size distribution, and nanoparticle fragmentation and/or aggregation.  RiboGreen RNA assay was used to determine siRNA encapsulation and retention.

Results: DMSO is ineffective as a cryoprotectant and disrupts liposome membranes leading to siRNA leakage and lysis after 24h storage at all temps. DLS data for DMEM and BSA suggested that morphology was maintained for up to 48hr, but siRNA retention was not improved at any temp.  CPP-PLPs stored in PBS exhibited the best siRNA retention after 24 and 48hr storage at 4°C, while storage at -20°C resulted in more siRNA leakage.

Conclusion: Our data suggests the optimal parameters for short-term CPP-PLP/siRNA complex storage are in PBS at 4°C while DMSO, DMEM, and BSA were ineffective at maintaining liposome stability.  Further experiments are ongoing to determine siRNA retention under these conditions over time, and how lipid modifications can increase stability for more efficient long-term storage.  These studies will direct our experimental implementation of this novel drug delivery technology in future in vivo applications in vascular injury models of disease.   

 

42.07 Depression Severity and Increased Inflammation in Veterans with Peripheral Artery Disease

Posted on January 18, 2018

N. V. Hernandez1, S. Grenon1,2, J. L. Ramirez1, S. A. Khetani1,2, W. J. Gasper1,2, D. Lindqvist3,4, O. M. Wolkowitz3, N. K. Hills5, G. J. Zahner1  2Veterans Affairs Medical Center,Vascular Surgery Section,San Francisco, CA, USA 3University Of California – San Francisco,Department Of Psychiatry,San Francisco, CA, USA 4Lund University,Department Of Psychiatry,Lund, , Sweden 5University Of California – San Francisco,Department Of Epidemiology And Biostatistics,San Francisco, CA, USA 1University Of California – San Francisco,Department Of Surgery,San Francisco, CA, USA

Introduction:

Inflammation is an important factor in the pathogenesis of peripheral artery disease (PAD) and is known to play a role in depression. The present study seeks to understand the association between depressive symptoms and inflammation in patients with PAD.

Methods:

A cross-sectional sample of 117 patients with PAD (97% male and 76% Caucasian) was recruited from the San Francisco Veterans Affairs Medical Center. Inclusion criteria included claudication symptoms with an abnormal ankle-brachial index (ABI < 0.9) or a history of prior peripheral revascularization for symptomatic PAD. Patients taking immunosuppressants and those with renal, liver, or other acute disease states were excluded. Patients were categorized into three subgroups based upon worsening depressive symptoms defined by Patient Health Questionnaire (PHQ-9) scores: no symptoms (score of 0-4, n=62), mild symptoms (score of 5-9, n=33), and depression (score ≥10, n=22). Serum levels of high-sensitivity C-reactive protein (CRP), interleukin-6 (IL-6), soluble intercellular adhesion molecule-1 (ICAM-1), and tumor necrosis factor-alpha (TNF-α ) were assayed and log-transformed for multivariable analysis. Additionally, inflammatory markers were standardized and summed to create a total inflammatory score.

Results:

Patients with depression exhibited a greater comorbidity with PTSD compared to patients with mild symptoms or no symptoms (57% vs 21% and 12% respectively, p≤.001) and were more likely to be Caucasian (95% vs 85% and 65% respectively, p=.02). In unadjusted analysis, inflammation levels trended upward among groups with no symptoms, mild symptoms, and depression, reported in respective order: IL-6 (medians: 1.17, 1.30, 1.65 IQRs: 0.93-1.73, 0.89-1.88, 1.20-2.41 pg/mL, p=.10), CRP (medians: 2.30, 3.65, 4.0 IQRs: 1.4-4.1, 1.75-6.0, 1.9-8.5 mg/L, p=.06), and total inflammatory score (medians: -1.1, -0.33, 1.3 IQRs: -1.88-0.77, -1.64-0.86, -1.07-3.21, p=.06). In a multivariable analysis that controlled for age, ethnicity, ABI, and atherosclerotic risk factors, mild symptoms and depression were predictive of a higher total inflammatory score (mild symptoms vs. no symptoms p=.04, and depression vs. no symptoms p=.007). IL-6 was marginally higher in the mild symptoms group (p=.06) and significantly higher in the depression group (p=.006) compared to the group with no symptoms. CRP, ICAM-1, and TNF-α  levels all trended upward among groups with depressive symptoms, but did not reach statistical significance.

Conclusions:

After controlling for risk factors and PAD severity, depression severity was independently associated with greater inflammation in PAD, as demonstrated by increasing levels of the total inflammatory score and of IL-6. Future research should examine the strength and directionality of this relationship through larger prospective cohort studies as well as investigate the pathophysiological mechanism responsible.

« Previous Page
Next Page »