Is Pulmonary Embolism Predictable in Patients With Deep Vein Thrombosis?

806 Abstracts

JOURNAL OF VASCULAR SURGERY September 2015

Inferior Vena Cava Filter Placement and Retrieval: A Survey of Vascular Specialists Practices Anand Brahmandam, MD1, Laura Skrip, MPH2, Bauer Sumpio, MD, PhD1, Jeffrey Indes, MD1, Alan Dardik, MD, PhD1, Timur Sarac, MD1, John Rectenwald, MD3, Cassius Iyad Ochoa Chaar, MD, MS1. 1Yale School of Medicine, New Haven, Conn; 2Yale School of Public Health, New Haven, Conn; 3University of Michigan School of Medicine, Ann Arbor, Mich

Fig. Receiver operator characteristic (ROC) curve analysis using common femoral vein (CFV) to predict >50% stenosis by intravascular ultrasound (IVUS).

Conclusions: The current study highlights the validity of VDUS in diagnosing iliac vein stenosis. Stent placement and pentoxyphylline were found to reduce the risk of recurrence. Author Disclosures: A. Mousa: None; M. Broce: None; M. Yacoub: None; J. Kazil: None; S. AbuHaliamh: None; A. Nanjundappa: None; P. Stone: None; M. Bates: None; A. AbuRahma: None. National Incidence and Ten-Year Trends in Deep Vein Thrombosis Following Total Knee and Total Hip Replacement Anahita Dua, MD1, Sapan S. Desai2, Jennifer A. Heller, MD3. 1Medical College of Wisconsin, Milwaukee, Wisc; 2Southern Illinois University, Springfield, Ill; 3Johns Hopkins University, Baltimore, Md Objectives: Total knee replacement (TKR) and total hip replacement (THR) are associated with an increased risk of deep vein thrombosis (DVT). Advances in DVT prophylaxis over the past decade have led to a decrease in DVT-related morbidity, but gender, racial, and other demographic factors that contribute to a higher risk of DVT are incompletely characterized. This study aimed to determine the incidence of DVT over the past decade and identify factors that were associated with an increased risk of DVT. Methods: Patients who underwent TKR or THR between 2001 and 2011 were identified using the National Inpatient Sample. Demographics (including age, gender, and race), comorbidities, and subsequent outcomes (including length of stay [LOS] and mortality) were determined for patients who developed a DVT. Differences between patients who did and did not develop a DVT were determined using multivariate regression analysis. A Mann-Kendall analysis was done to evaluate all trends. Results: Between 2001 and 2011, 1.1 million patients underwent TKR, and 550,000 underwent THR. The overall incidence of DVT decreased for TKR from 0.86% in 2001 to 0.45% in 2011 and decreased for THR from 0.55% to 0.24% for the same period. Patients who developed a DVT after TKR were older (67.7 vs 66.8 years; P < .001) and were more likely to be male (37.9% vs 36.4%; P < .001), African American (P < .001), and to have significant comorbidities, including congestive heart failure, peripheral artery disease, and end-stage renal disease. Findings were similar for patients who developed a DVT after THR. Mortality was significantly greater for patients who developed a DVT (0.4% after TKR and 1.7% after THR), with an almost double LOS. Conclusions: A focus on DVT prophylaxis has decreased national rates of DVT after TKR and THR. However, older patients, men, African Americans, and patients with more comorbidities appear to be especially at risk for DVT. Mortality was almost eight times higher than in patients who did not develop DVT, and a LOS double that of unaffected patients significantly affects the quality of care. A focus on DVT prophylaxis, and perhaps more aggressive management of the at-risk population, may help decrease the rate of DVT. Author Disclosures: A. Dua: None; S. S. Desai: None; J. A. Heller: None.

Objectives: The frequency of inferior vena cava filter (IVCF) placement continues to rise. Vascular specialists (VSs) adopt different practices based on local expertise. This study was performed to assess the attitudes of VSs toward the placement and retrieval of IVCFs. Methods: An online survey of 28 questions related to practice patterns regarding IVCFs was administered to a group of vascular surgeons and interventional radiologists. VSs were categorized as low volume (LV) if they place fewer than three IVCFs per month and as high volume (HV) if they place at least 3 IVCF per month. The responses of HV and LV were compared and analyzed using c2 tests. Results: A total of 259 VSs completed the survey. There were 199 vascular surgeons (74%) and 69 interventional radiologists (26%). Most responders (67%) were in academic practice and worked in tertiary care centers (73%). The temporary IVCF of choice was Celect (27%), followed by Denali (20%). The permanent IVCF of choice was a temporary IVCF left in situ indefinitely (42%), followed by Titanium Greenfield (17%). Eightytwo percent preferred placing the tip of the IVCF at or just below the lowest renal vein. Thirty-one percent obtained a venous duplex of the lower extremities before retrieval, and 34% did not do any imaging. VSs were divided into 132 LV (51%) and 127 HV (49%). HV responders were significantly more likely to have procedural time for IVCF retrieval of <30 minutes compared with LV responders (57% vs 42%; P ¼ .026). There was a trend for HV responders to have fewer unsuccessful attempts at IVCF retrieval but that did not reach statistical significance (P ¼ .061). HV responders were more likely to have attempted multiple times to retrieve an IVCF (66% vs 33%; P < .001) and to have used a bronchoscopy forceps (32% vs 14%; P ¼ .001) or a laser sheath (14% vs 2%; P < .001) for IVCF retrieval. In general, VSs were not comfortable using a bronchoscopy forceps (65%) or a laser sheath (90%) for IVCF retrieval. Conclusions: This study underscores significant variability in VS practice patterns regarding IVCFs. More studies and societal guidelines are needed to define the best practices. Author Disclosures: A. Brahmandam: None; L. Skrip: None; B. Sumpio: None; J. Indes: None; A. Dardik: None; T. Sarac: None; J. Rectenwald: None; C. Ochoa Chaar: None. Is Pulmonary Embolism Predictable in Patients With Deep Vein Thrombosis? Nancy Huynh, BS, Wassim Fares, MD, MPH, Kirstyn Brownson, MD, Anand Brahmandam, MD, Alfred I. Lee, MD, PhD, Alan Dardik, MD, PhD, Timur Sarac, MD, Cassius Iyad Ochoa Chaar. Yale School of Medicine, New Haven, Conn Objectives: The Caprini model estimates patients risk for venous thromboembolism (VTE) based on 30 different factors. Hemodynamically significant (HS) pulmonary embolism (PE), defined as high risk (massive) or intermediate (submassive) PE, is the most dreaded complication. This study tests whether the Caprini model correlates with the prevalence of PE and HS PE in patients with deep vein thrombosis (DVT). Methods: A retrospective review of the records of all consecutive patients diagnosed with DVT between January 2013 and August 2014 in a tertiary care center was performed. Patient demographics and risk factors based on the Caprini model were noted. Multivariable analysis was used to determine predictors of PE, and HS PE. Stata software was used for all analysis. Results: There were 838 patients (50.95% women) with DVT; 217 patients (25.89%) had a concomitant PE at presentation and 135 had HS PE (101 submassive PE, 34 massive PE). The mean age was 65 years. There was no significant relation between age or gender and the occurrence of PE or HS PE. Patients with PE were less likely to have undergone major surgery (18.43% vs 81.57%; P ¼ .001), to have inflammatory bowel disease (3.69% vs 96.31%; P ¼ .001), and to have sepsis (4.61% vs 95.39%; P ¼ .002) but were more likely to be overweight (body mass index >25 kg/ m2; 74.65% vs 24.42%; P ¼ .007). On multivariable analysis, only major surgery (P ¼ .001) and sepsis (P ¼ .008) remained statistically significant. The Caprini score had statistically significant inverse relation with occurrence of PE, in that patients with a higher score were less likely to develop PE (P ¼ .047). Patients with DVT after major surgery were less likely to have HS PE (13.33% vs 86.67%; P ¼ .011). There was no association between Caprini score and HS PE (P ¼ .171).

JOURNAL OF VASCULAR SURGERY Volume 62, Number 3

Abstracts 807

Conclusions: The Caprini model has a poor correlation with PE or HS PE in patients with DVT. Among all patients with DVT, a concomitant diagnosis of PE or HSPE is less common in those undergoing major surgery. Author Disclosures: N. Huynh: None; W. Fares: None; K. Brownson: None; A. Brahmandam: None; A. I. Lee: None; A. Dardik: None; T. Sarac: None; C. Ochoa Chaar: None. Effect of Iliac Vein Stenting of NIVLs on Venous Reflux Times Yuriy Ostrozhynskyy, Anil Hingorani, MD, Enrico Ascher, MD, Eleanor Iadgarova, RN, Arkady Ganelin, MD, Natalie Marks, MD. Total Vascular Care, Brooklyn, NY Objectives: Iliac vein stenting of nonthrombotic iliac vein lesions (NIVL) is becoming a common treatment option for venous insufficiency. This treatment option helps relieve symptoms of venous insufficiency that is derived from venous reflux. We examined the effect that iliac vein stenting had on reflux times before and after stenting. Methods: Over the course of 22 months, we performed 655 venograms with angioplasties and stenting of the iliac veins. These procedures were filtered to only include the venograms where a stent was placed, and no other vascular procedures was performed between “pre” and “post” venous mapping tests. This was done to ensure that the iliac vein stent placed was the only procedure affecting the venous reflux. This left 273 procedures performed over the course of 14 months. Reflux was measured using ipsilateral ultrasound. Results: The patients were an average age of 70.44 years (standard deviation, 14.12; range, 23-95 years), with 177 women and 96 men. The CEAP scores were C2 in 11%, C3 in 53%, C4 in 11%, C5 in 3%, and C6 in 22%. Table reports the average reflux values before and after stenting and the tendency of the decrease in reflux. Conclusions: These data reveal a uniform decrease in the vein reflux of prestent and poststent reflux values. However, no statistical significance was calculated between the prestent and poststent average reflux time, probably due to the large standard deviation. Table. Prestent and poststent average vein reflux in milliseconds Laterality

Vein segment

RLE

Common femoral Femoral Popliteal GSV SSV Common femoral Femoral Popliteal GSV SSV

RLE RLE RLE RLE LLE LLE LLE LLE LLE

Prestent

SD

411.2

561.9

782.1 964.5 2157 1811.9 356.6 470.5 1189.9 2173 1856.3

Poststent

SD

P value

66.6

230.7

.4362

1077 848.2 1586.5 1522.6 418

507.2 683.8 1856.6 1323.3 53.6

667.3 938.6 1772.7 1321.2 177.9

.4835 .289 .4807 .2087 .3481

619.3 1067.9 1617 1452.4

408.3 960.5 2064.4 1402.9

812.5 1225.2 1694.4 1268.2

.8718 .847 .7263 .1567

GSV, Great saphenous vein; LLE, left lower extremity; RLE, right lower extremity; SD, standard deviation; SSV, short saphenous vein. Author Disclosures: Y. Ostrozhynskyy: None; A. Hingorani: None; E. Ascher: None; E. Iadgarova: None; A. Ganelin: None; N. Marks: None. Endoscopic Vein Harvest (EVH) Does Not Negatively Impact Patency of Great Saphenous Vein (GSV) Lower Extremity Bypass Sikandar Khan1, Mariel Rivero, MD2, Brian McCraith3, Linda Harris, MD1, Maciej Dryjski, MD1, Hasan H. Dosluoglu2. 1SUNY at Buffalo, Buffalo, NY; 2VA Western NY Healthcare System, SUNY at Buffalo, Buffalo, NY; 3 VA Western NY Healthcare System, Buffalo, NY Objectives: Endoscopic vein harvest (EVH) has been reported to reduce the morbidity and length of stay compared with open vein harvest (OVH) for infrainguinal bypass procedures, but there have been concerns regarding decreased graft patency and increased rates of reinterventions with EVH compared with OVH. We have been using EVH since 2007 in collaboration with our cardiac surgery team, and this is our currently preferred approach. The goal of this study was to see if EVH is comparable to OVH in graft patency and limb salvage and associated with fewer wound complications.

Methods: Included were 175 patients (193 limbs) undergoing elective lower extremity bypass with single-piece autologous GSV from June 1, 2001, to December 31, 2014. Patients were followed up postoperatively and every 6 months clinically and with venous duplex. Patency, limb salvage rates, and perioperative complications were compared between OVH and EVH. Results: Seventy-four patients (85 limbs) had EVH, and 101 (108 limbs) had OVH, and 78% and 84%, respectively, had critical limb ischemia (P ¼ .242). Comorbidities were similar, but more OVH patients were on warfarin (38% vs 21%; P ¼ .012) and more EVH patients were on ECASA (93% vs 73%; P < .001) and on Plavix (64% vs 43%). Mean vein diameter was 3.2 6 0.7 mm. Distal targets were similar (6% above-knee, 25% below-knee, and 69% infrapopliteal). Mean length of stay was similar in EVH (8.3 6 8.6) and OVH (9.9 6 11.7) patients (P ¼ .095), but major wound complications were significantly less in the EVH group (8% vs 19%; P ¼ .012). Mean follow-up was 38 6 34 months. Primary patency at 5 years (62% 6 8% EVH vs 56% 6 6% OVH; P ¼ .983), assisted primary (79% 6 6% EVH vs 71% 6 6% OVH; P ¼ .865), secondary patency (86% 6 5% EVH vs 75% 6 5% OVH; P ¼ .409), limb salvage (83% 6 7% EVH vs 81% 6 4% OVH; P ¼ .217), freedom from major adverse limb events, major adverse cardiac events, RAO, and RAS were all similar between two groups. Conclusions: In experienced hands, EVH is associated with significant decrease in wound complications with similar graft patency, reintervention rates and limb salvage. Author Disclosures: S. Khan: None; M. Rivero: None; B. McCraith: None; L. Harris: None; M. Dryjski: None; H. H. Dosluoglu: None. Equivalent Outcomes With Standard and Heparin-Bonded Expanded Polytetrafluoroethylene Grafts Used as Conduits for Hemodialysis Access Mark G. Davies, MD, PhD1, Hosam F. ElSayed, MD2. 1University of Texas Health Science Center at San Antonio, San Antonio, Tex; 2University of Texas Health Science Center at San Antonio, Ohio State University, Ohio Objectives: Obtaining and maintaining dialysis access after failure of autologous access sites remains a significant concern. Polytetrafluoroethylene (PTFE) is the most common conduit used. Recently, heparin-bonded expanded PTFE (HB-PTFE) grafts have been introduced as improved conduits with suggestions that they offer benefits because of their resistance to thrombosis due to the heparin bonding. This retrospective study compares the outcomes of HB-PTFE compared with standard PTFE (S-PTFE) arteriovenous grafts (AVG). Methods: From January 2004 to December 2014, 483 adult patients (46% male; mean age, 60; range, 25-87 years) with end-stage renal disease underwent placement of AVGs (234 HB-PTFE and 248 S-PTFE). There were no significant differences in demographics or access history between the two groups. Patency, reintervention, infection, and functional dialysis rates were examined and factors influencing their outcomes determined. Results: Technical success was 99% in HB-PTFE and 98% in S-PTFE. The 30-day major adverse cardiac events were 2% in HB-PTFE and 3% in SPTFE. Average time to access was 5 6 0.8 weeks for HB-PTFE and 7 6 0.9 weeks for S-PTFE. Median follow-up was 23 months. The 2-year primary, assisted primary, and secondary patency rates were 20% 6 7% vs 18% 6 8% (P ¼ .85), 35% 6 8% vs 28% 6 7% (P ¼ .51), and 38% 6 6% vs 36% 6 7% (P ¼ .83) for HB-PTFE vs S-PTFE, respectively. The number of secondary interventions was similar in the both groups (2.1 and 1.9 interventions/person-year of follow-up for HB-PTFE vs S-PTFE, respectively; P ¼ .87). There were no significant differences in infection (11% vs 12%) or pseudoaneurysm (5% vs 6%) between HB-PTFE and S-PTFE groups. Functional dialysis durations were equivalent between HB-PTFE and S-PTFE groups. Conclusions: HB-PTFE grafts offer no distinct advantage over SPTFE grafts for hemodialysis and should not be considered a preferential conduit for arteriovenous grafting. Author Disclosures: M. G. Davies: None; H. F. ElSayed: None. Novel Strong Association Between Bone Mineral Density Scores and the Prevalence of Peripheral Arterial Disease in Both Sexes Caron B. Rockman, MD1, Thomas Maldonado, MD1, Jade Hiramoto, MD2, Stephen Honig, MD1, Michael Conte, MD2, Jeffrey Berger, MD1. 1New York University Medical Center, New York, NY; 2 University of California, San Francisco, New York, NY Objectives: There is a reported association between bone mineral density (BMD) scores and coronary disease. The association of peripheral arterial disease (PAD) and BMD scores has not been extensively investigated. Our objective was to investigate the association between BMD scores