Adverse Technical Events During Aortic Root Replacement.

BACKGROUND: Although adverse technical events during aortic root replacement (ARR) are not uncommon and are extremely challenging, there is little literature to help surgeons prepare for such situations. We describe our experience of outstanding technical events during ARR. METHODS: This is a retrospective study of 830 consecutive ARR at a single center from 2012-2022. Technical events were defined as intraoperative events that led to an unplanned cardiac procedure, need for mechanical circulatory support, or additional aortic cross-clamping. Logistic regression identified factors associated with operative mortality and technical events. RESULTS: Technical events occurred in 90 (10.8%) patients, and were attributed to bleeding (n=26), nonischemic ventricular dysfunction (n=23), residual valve disease (n=20), myocardial ischemia (n=19), and iatrogenic dissection (n=2). Prior sternotomy (OR 2.38 [1.36-4.19], p=0.002) and complex aortic valve disease (OR 3.09 [1.09-8.75], p=0.03) were associated with technical events. Patients with technical events had higher rates of operative mortality (6.7% vs 2.3%, p=0.03) and all major postoperative complications. Surgical indications of dissection (OR 13.57 [4.95-37.23], p<0.001) and complex aortic valve disease (OR 14.09 [3.67-54.02], p<0.001) but not adverse technical events (OR 2.42 [0.81-7.26], p=0.11) were associated with operative mortality. CONCLUSIONS: Adverse technical events occur in 10.8% of ARR and were associated with reoperative sternotomies. Technical events are associated with increased postoperative complications.

Mechanical Circulatory Support During Surgical Revascularization for Ischemic Cardiomyopathy.

BACKGROUND: This study aimed to describe the use of perioperative mechanical circulatory support (MCS) and its impact on outcomes in patients with ischemic cardiomyopathy who were undergoing surgical revascularization. METHODS: Patients with an ejection fraction <35% who underwent isolated coronary artery bypass grafting (CABG) from 2015 to 2021 were identified (N = 378). Patients were divided into no MCS, preoperative MCS, and postoperative MCS groups on the basis of timing of MCS initiation, which included intraaortic balloon pump, extracorporeal membrane oxygenation, or Impella device (Abiomed) use. The primary outcome of interest was operative mortality. RESULTS: The median Society of Thoracic Surgeons Predicted Risk of Mortality was 2.4%. Sixty-six percent (n = 246) of patients had a previous myocardial infarction, and 61.8% of these patients were within 21 days of CABG. Twenty-one patients (5.6%) presented in cardiogenic shock. The preoperative MCS cohort consisted of 31 patients (8.2%) who underwent CABG a median of 2 days after MCS initiation. Thirty (7.9%) patients required postoperative MCS. Independent risk factors for requiring postoperative MCS included the preoperative ejection fraction (odds ratio, 0.93; P = .01 and the presence of preoperative MCS (odds ratio, 3.06; P = .02). Overall, operative mortality was 3.4%, and 3-year survival was 87.0%. Operative mortality in patients who did and did not receive preoperative MCS was 7.7% and 2.9% (P = .12) with no difference in long-term survival (P = .80), whereas patients requiring postoperative MCS had significantly increased operative (16.7%) and late mortality (63%; P <.01). CONCLUSIONS: CABG can be performed safely in patients with ischemic cardiomyopathy with selective use of perioperative MCS. Despite advanced disease severity, patients requiring preoperative MCS demonstrate acceptable short- and long-term survival. Patients requiring postoperative MCS have increased postoperative morbidity and mortality.

Hemodynamic Optimization by Invasive Ramp Test in Patients Supported With HeartMate 3 Left Ventricular Assist Device.

In patients supported by the HeartMate 3 left ventricular assist device (HM3 LVAD), pump speed adjustments may improve hemodynamics. We investigated the hemodynamic implications of speed adjustments in HM3 recipients undergoing hemodynamic ramp tests. Clinically stable HM3 recipients who underwent routine invasive hemodynamic ramp tests between 2015 and 2022 at our center were included. Filling pressure optimization, defined as central venous pressure (CVP) <12 mm Hg and pulmonary capillary wedge pressure (PCWP) <18 mm Hg, was assessed at baseline and final pump speeds. Patients with optimized pressures were compared to nonoptimized patients. Overall 60 HM3 recipients with a median age of 62 years (56, 71) and time from LVAD implantation of 187 days (124, 476) were included. Optimized filling pressures were found in 35 patients (58%) at baseline speed. Speed was adjusted in 84% of the nonoptimized patients. Consequently, 39 patients (65%) had optimized pressures at final speed. There were no significant differences in hemodynamic findings between baseline and final speeds (p > 0.05 for all). Six and 12 month readmission-free rates were higher in optimized compared with nonoptimized patients (p = 0.03 for both), predominantly due to lower cardiac readmission-free rates (p = 0.052). In stable outpatients supported with HM3 who underwent routine ramp tests, optimized hemodynamics were achieved in only 2 of 3 of the patients. Patients with optimized pressures had lower all-cause readmission rates, primarily driven by fewer cardiac-related hospitalizations.

The Hemodynamic Effects of Aortic Regurgitation in Patients Supported by a HeartMate 3 Left Ventricular Assist Device.

BACKGROUND: Aortic regurgitation (AR) is a common complication following left ventricular assist device (LVAD) implantation. We evaluated the hemodynamic implications of AR in patients with HeartMate 3 (HM3) LVAD at baseline and in response to speed changes. METHODS AND RESULTS: Clinically stable outpatients supported by HM3 who underwent a routine hemodynamic ramp test were retrospectively enrolled in this analysis. Patients were stratified based on the presence of at least mild AR at baseline speed. Hemodynamic and echocardiographic parameters were compared between the AR and non-AR groups. Sixty-two patients were identified. At the baseline LVAD speed, 29 patients (47%) had AR, while 33 patients (53%) did not. Patients with AR were older and supported on HM3 for a longer duration. At baseline speed, all hemodynamic parameters were similar between the groups including central venous pressure, pulmonary capillary wedge pressure, pulmonary arterial pressures, cardiac output and index, and pulmonary artery pulsatility index (p > 0.05 for all). During the subacute assessment, AR worsened in some, but not all, patients, with increases in LVAD speed. There were no significant differences in 1-year mortality or hospitalization rates between the groups, however, at 1-year, ≥ moderate AR and right ventricular failure (RVF) were detected in higher rates among the AR group compared to the non-AR group (45% vs. 0%; p < 0.01, and 75% vs. 36.8%; p = 0.02, respectively). CONCLUSIONS: In a cohort of stable outpatients supported with HM3 who underwent a routine hemodynamic ramp test, the presence of mild or greater AR did not impact the ability of HM3 LVADs to effectively unload the left ventricle during early subacute assessment. Although the presence of AR did not affect mortality and hospitalization rates, it resulted in higher rates of late hemodynamic-related events in the form of progressive AR and RVF.

Comparing 3-year survival and readmissions between HeartMate 3 and heart transplant as primary treatment for advanced heart failure.

OBJECTIVE: To compare 3-year survival and readmissions of patients who received the HeartMate 3 (HM3) left ventricular assist device (LVAD) or underwent orthotopic heart transplantation (OHT) as primary treatment for advanced heart failure. METHODS: We retrospectively analyzed 381 adult patients who received an HM3 LVAD or were listed for OHT between January 2014 and March 2021 at our center. To minimize crossover bias, OHT recipients with a prior LVAD were excluded, and HM3 patients were censored at the time of transplant. Cohorts were propensity score-matched to reduce confounding variables. The primary outcome was 3-year survival, and the secondary outcome was mean cumulative all-cause unplanned readmission. RESULTS: The study population comprised 185 HM3 patients (49%) and 196 OHT patients (51%), with 104 propensity score-matched patients in each group. After propensity score matching, there was no statistical difference in 3-year survival (83.7% for HM3 vs 87.0% for OHT; P = .91; relative risk [RR], 1.00; 95% confidence interval [CI], 0.45-2.20). In the unmatched cohorts, patients age 18 to 49 years had comparable survival with HM3 and OHT (96.9% vs 95.9%; N = 91; P = 1.00; RR, 0.92; 95% CI, 0.09-9.78). Patients age 50+ years had slightly inferior survival with HM3 (75.0% vs 83.9%; N = 290; P = .60; RR, 1.51; 95% CI, 0.85-2.68). The mean number of readmissions at 3 years was higher in the HM3 group (3.89 vs 2.05; P < .001). CONCLUSIONS: This exploratory analysis suggests that for similar patients, HM3 may provide comparable 3-year survival to OHT as a primary treatment for heart failure but may result in more readmissions.

Aortic Root Thrombosis in patients with HeartMate 3 left ventricular assist device support.

BACKGROUND: Aortic root thrombosis(ART) is a complication of continuous-flow left ventricular assist device therapy. However, the incidence and related complications of ART in HeartMate 3 (HM3) patients remain unknown. METHODS: Patients who underwent HM3 implantation from November 2014 to August 2020 at a quaternary academic medical center were included. Demographics and outcomes were abstracted from the medical record. Echocardiograms and contrast-enhanced computed tomography studies were reviewed to identify patients who developed ART and/or moderate or greater aortic insufficiency (AI) on HM3 support. RESULTS: The study cohort included 197 HM3 patients with a median postimplant follow-up of 17.5 months. Nineteen patients (9.6%) developed ART during HM3 support, and 15 patients (7.6%) developed moderate or greater AI. Baseline age, gender, race, implantation strategy, and INTERMACS classification were similar between the ART and no-ART groups. ART was associated with an increased risk of death, stroke, or aortic valve (AV) intervention (subhazard ratio [SHR] 3.60 [95% confidence interval (CI) 1.71-7.56]; p = 0.001) and moderate or greater AI (SHR 11.1 [CI 3.60-34.1]; p < 0.001) but was not associated with a statistically significantly increased risk of death or stroke on HM3 support (2.12 [0.86-5.22]; p = 0.10). Of the 19 patients with ART, 6 (31.6%) developed moderate or greater AI, necessitating more frequent AV interventions (ART: 5 AV interventions [3 surgical repairs, 1 surgical replacement, 1 transcatheter replacement; 26.3%]; no-ART: 0). CONCLUSIONS: Nearly 10% of HM3 patients developed ART during device support. ART was associated with increased risk of a composite end-point of death, stroke, or AV intervention as well as moderate or greater AI.

Effect of venoarterial extracorporeal membrane oxygenation initiation timing on tricuspid valve surgery outcomes.

Objectives: Tricuspid valve surgery is associated with high rates of shock and in-hospital mortality. Early initiation of venoarterial extracorporeal membrane oxygenation after surgery may provide right ventricular support and improve survival. We evaluated mortality in patients undergoing tricuspid valve surgery based on the timing of venoarterial extracorporeal membrane oxygenation. Methods: All consecutive adult patients undergoing isolated or combined surgical tricuspid valve repair or replacement from 2010 to 2022 requiring venoarterial extracorporeal membrane oxygenation use were stratified by initiation in the operating room (Early) versus outside of the operating room (Late). Variables associated with in-hospital mortality were explored using logistic regression. Results: There were 47 patients who required venoarterial extracorporeal membrane oxygenation: 31 Early and 16 Late. Mean age was 55.6 years (standard deviation, 16.8), 25 (54.3%) were in New York Heart Association class III/IV, 30 (60.8%) had left-sided valve disease, and 11 (23.4%) had undergone prior cardiac surgery. Median left ventricular ejection fraction was 60.0% (interquartile range, 45-65), right ventricular size was moderately to severely increased in 26 patients (60.5%), and right ventricular function was moderately to severely reduced in 24 patients (51.1%). Concomitant left-sided valve surgery was performed in 25 patients (53.2%). There were no differences in baseline characteristics or invasive measurements immediately before surgery between the Early and Late groups. Venoarterial extracorporeal membrane oxygenation was initiated 194 (23.0-840.0) minutes after cardiopulmonary bypass in the Late venoarterial extracorporeal membrane oxygenation group. In-hospital mortality was 35.5% (n = 11) in the Early group versus 68.8% (n = 11) in the Late group (P = .037). Late venoarterial extracorporeal membrane oxygenation was associated with in-hospital mortality (odds ratio, 4.00; 1.10-14.50; P = .035). Conclusions: Early postoperative initiation of venoarterial extracorporeal membrane oxygenation after tricuspid valve surgery in high-risk patients may be associated with improvement in postoperative hemodynamics and in-hospital mortality.

Society for Cardiovascular Angiography and Interventions Shock Classification to Stratify Outcomes of Extracorporeal Membrane Oxygenation.

We applied the Society for Cardiovascular Angiography and Interventions (SCAI) schema to cardiogenic shock (CS) patients treated with venoarterial extracorporeal membrane oxygenation (VA-ECMO) to assess performance in this high acuity group of patients. Records of adult patients receiving VA-ECMO for CS at our institution from 01/2015 to 12/2019 were reviewed. Post-cardiotomy and noncardiogenic shock patients were excluded. A total of 245 patients were included, with a median age of 59 years [IQR: 48-67]; 159 (65%) were male. There were 34 (14%) patients in Stage C, 82 (33%) in D, and 129 (53%) in E. Of E patients, 88 (68%) were undergoing cardiopulmonary resuscitation. Median ECMO duration decreased with stage (C:7, D:6, E:4 days, P < 0.001). In-hospital mortality increased (C:35%, D:56%, E:71%, P < 0.001) and myocardial recovery decreased with stage (C:65%, D:35%, E:30%, P < 0.001). Acute kidney injury (C:35%, D:45%, E:54%, P = 0.045), acute liver failure (C:32%, D:66%, E:76%, P < 0.001), and infection (C:35%, D:28%, E:16%, P = 0.004) varied among groups. Multivariable analysis revealed age (HR=1.02), male sex (HR=0.62), and E classification (HR=2.69) as independently associated with 1-year mortality. Competing-risks regression identified D (SHR=0.53) and E classification (SHR=0.45) as inversely associated with myocardial recovery. In patients treated with VA-ECMO for CS, the SCAI classification provided robust risk stratification.

Continuous near-infrared reflectance spectroscopy monitoring to guide distal perfusion can minimize limb ischemia surgery for patients requiring femoral venoarterial extracorporeal life support.

BACKGROUND: Patients requiring femoral venoarterial (VA) extracorporeal life support (ECLS) are at risk of distal lower limb hypoperfusion and ischemia of the cannulated leg. In the present study, we evaluated the effect of using continuous noninvasive lower limb oximetry with near-infrared reflectance spectroscopy (NIRS) to detect tissue hypoxia and guide distal perfusion catheter (DPC) placement on the rates of leg ischemia requiring surgical intervention. METHODS: We performed a retrospective analysis of patients who had undergone femoral VA-ECLS at our institution from 2010 to 2014 (pre-NIRS era) and 2017 to 2021 (NIRS era). Patients who had undergone cannulation during the 2015 to 2016 transition era were excluded. The baseline characteristics, short-term outcomes, and ischemic complications requiring surgical intervention (eg, fasciotomy, thrombectomy, amputation, exploration) were compared across the two cohorts. RESULTS: Of the 490 patients included in the present study, 141 (28.8%) and 349 (71.2%) had undergone cannulation before and after the routine use of NIRS to direct DPC placement, respectively. The patients in the NIRS cohort had had a greater incidence of hyperlipidemia (53.7% vs 41.1%; P = .015) and hypertension (71.4% vs 60%; P = .020) at baseline, although they were less likely to have been supported with an intra-aortic balloon pump before ECLS cannulation (26.9% vs 37.6%; P = .026). These patients were also more likely to have experienced cardiac arrest (22.9% vs 7.8%; P ≤ .001) and a pulmonary cause (5.2% vs 0.7%; P = .04) as an indication for ECLS, with ECLS initiated less often for acute myocardial infarction (15.8% vs 34%; P ≤ .001). The patients in the NIRS cohort had had a smaller arterial cannula size (P ≤ .001) and a longer duration of ECLS support (5 vs 3.25 days; P ≤ .001) but significantly lower rates of surgical intervention for limb ischemia (2.6% vs 8.5%; P = .007) despite comparable rates of DPC placement (49.1% vs 44.7%; P = .427), with only two patients (1.1%) not identified by NIRS ultimately requiring surgical intervention. CONCLUSIONS: The use of a smaller arterial cannula (≤15F) and continuous NIRS monitoring to guide selective insertion of DPCs could be a valid and effective strategy associated with a reduced incidence of ischemic events requiring surgical intervention.

Outcomes following successful decannulation from extracorporeal life support for cardiogenic shock.

OBJECTIVE: Although extracorporeal life support (ECLS) has increasingly been used for the treatment of patients with cardiogenic shock (CS), the outcomes of those successfully weaned from support remain poorly defined. METHODS: Of 510 venoarterial ECLS CS patients at our institution between January 2015 and December 2020, 249 were decannulated and survived for 30 days or until discharge (ie, successfully weaned). Factors associated with survival to discharge were assessed and 1-year survival was described. RESULTS: Of 510 eligible CS ECLS patients, 249 (48.8%) were successfully decannulated, 227 (44.5%) died during/following ECLS, and 34 (6.7%) were bridged to heart transplantation or a ventricular assist device. Patients with a primary graft dysfunction etiology of CS had a greater chance of successful decannulation (odds ratio [OR], 3.088; 95% CI, 1.1-8.671; P = .0323), whereas patients with ECLS during cardiopulmonary resuscitation had a reduced chance of successful decannulation (OR, 0.354; 95% CI, 0.17-0.735; P = .0054). Of successfully decannulated patients, 218 (87.6%) survived to hospital discharge and 31 (12.4%) died in the hospital. Acute myocardial infarction etiology (OR, 4.751; 95% CI, 1.623-13.902; P = .0044), preexisting chronic kidney disease (OR, 3.422; 95% CI, 1.374-8.52; P = .0082), and initiation of continuous renal replacement therapies (OR, 3.188; 95% CI, 1.291-7.871; P = .012) were significantly associated with in-hospital mortality despite successful decannulation. One-year survival in successfully decannulated patients surviving to hospital discharge was 95.0% and comparable to 1-year survival in patients who received a heart transplant or ventricular assist device. CONCLUSIONS: Successful decannulation can be achieved in a significant proportion of patients treated with ECLS for CS but does not guarantee survival to hospital discharge. However, 1-year survival of hospital survivors remains high and is comparable to patients bridged to transplant or a ventricular assist device.

Assessing sex-based differences in postsurgical clinical outcomes after use of del Nido cardioplegia.

OBJECTIVES: del Nido cardioplegia is used to pharmacologically arrest the heart during cardiac surgery and decrease reperfusion- and ischaemia-related myocardial injury. Studies have demonstrated the physiological differences between male and female hearts, potentially related to cardiac size or myocyte calcium handling; we aimed to assess for between-sex differences in clinical outcomes after receipt of del Nido cardioplegia. METHODS: Patients who underwent coronary artery bypass or coronary artery bypass graft/valve surgery at our institution using del Nido cardioplegia (January 2014 to December 2019) were included (n = 2118). Clinical data were collected retrospectively. After the creation of a propensity-matched cohort (n = 1252), multivariable logistic regression was used to analyse binary postoperative outcomes, and a Gamma model was used for a continuous postoperative outcome. Our primary end-point was a composite end-point comprised of 30-day mortality and/or need for a post-bypass mechanical support device. RESULTS: The final cohort included 459 females and 793 males (matched up to 1:2, all standardized mean differences <0.1). Multivariable logistic regression showed that biological sex was not associated with the composite primary end-point (odds ratio = 0.898, P = 0.779). A Gamma model indicated that there were no sex-related differences in vasoactive-inotropic scores reflecting vasopressor and inotrope usage at the time of patient operating room exit (exp[est] = 1.394, P = 0.189). CONCLUSIONS: Our findings showed no significant between-sex differences in clinical outcomes after receiving del Nido cardioplegia, suggesting adequate myocardial protection as currently administered. Further research is warranted to elicit if there are sex-based differences between cardioplegic solutions. IRB APPROVAL DATE (PROTOCOL NUMBER): 26 May 2021 (AAAR8359).

Predictors of 1-year outcome after cardiac re-transplantation: Machine learning analysis.

BACKGROUND: As cardiac re-transplantation is associated with inferior outcomes compared with primary transplantation, allocating scarce resources to appropriate re-transplant candidates is important. The aim of this study is to elucidate the factors associated with 1-year mortality in cardiac re-transplantation using the random forests algorithm for survival analysis. METHODS: We retrospectively reviewed the United Network for Organ Sharing registry and identified all adult (> 17 years old) recipients who underwent cardiac re-transplantation between January 2000 and March 2020. The random forest algorithm on Cox modeling was used to calculate the variable importance (VIMP) of independent variables for contributing to 1-year mortality. RESULTS: A total of 1294 patients underwent cardiac re-transplantation. Of these, 137 patients were re-transplanted within 1 year of their first transplant, while 1157 patients were re-transplanted more than 1 year after their first transplant. One-year mortality was significantly higher for patients receiving early transplantation compared with those receiving late transplantation (Early 40.6% vs. Late 13.6%, log-rank P < .001). Machine learning analysis showed that total bilirubin (> 2 mg/dl) (VIMP, 2.99%) was an independent predictor of 1-year mortality after early re-transplant. High BMI (> 30.0 kg/m2 ) (VIMP, 1.43%) and ventilator dependence (VIMP, 1.47%) were independent predictors of 1-year mortality for the late re-transplantation group. CONCLUSION: Machine learning showed that optimal 1-year survival following cardiac re-transplantation was significantly related to liver function in early re-transplantation, and to obesity and preoperative ventilator dependence in late re-transplantation.

Comparison of Risk Models in the Prediction of 30-Day Mortality in Acute Myocardial Infarction-Associated Cardiogenic Shock.

Background: There are numerous risk-prediction models applied to acute myocardial infarction-related cardiogenic shock (AMI-CS) patients to determine a more accurate prognosis and to assist in patient triage. There is wide heterogeneity among the risk models including the nature of predictors evaluated and their specific outcome measures. The aim of this analysis was to evaluate the performance of 20 risk-prediction models in AMI-CS patients. Methods: Patients included in our analysis were admitted to a tertiary care cardiac intensive care unit with AMI-CS. Twenty risk-prediction models were computed utilizing vitals assessments, laboratory investigations, hemodynamic markers, and vasopressor, inotropic and mechanical circulatory support available from within the first 24 â€‹hours of presentation. Receiver operating characteristic curves were used to assess the prediction of 30-day mortality. Calibration was assessed with a Hosmer-Lemeshow test. Results: Seventy patients (median age 63 years, 67% male) were admitted between 2017 and 2021. The models' area under the curve (AUC) ranged from 0.49 to 0.79, with the Simplified Acute Physiology Score II score having the most optimal discrimination of 30-day mortality (AUC: 0.79, 95% confidence interval [CI]: 0.67-0.90), followed by the Acute Physiology and Chronic Health Evaluation-III score (AUC: 0.72, 95% CI: 0.59-0.84) and the Acute Physiology and Chronic Health Evaluation-II score (AUC: 0.67, 95% CI: 0.55-0.80). All 20 risk scores demonstrated adequate calibration (p > 0.05 for all). Conclusions: Among the models tested in a data set of patients admitted with AMI-CS, the Simplified Acute Physiology Score II risk score model demonstrated the highest prognostic accuracy. Further investigations are required to improve the discriminative capabilities of these models or to establish new, more streamlined and accurate methods for mortality prognostication in AMI-CS.

Sex differences in patients with cardiogenic shock requiring extracorporeal membrane oxygenation.

OBJECTIVE: Our study assesses differences between male and female patients placed on venoarterial extracorporeal membrane oxygenation for cardiogenic shock. METHOD: We retrospectively analyzed 574 adult patients placed on venoarterial extracorporeal membrane oxygenation for cardiogenic shock at our institution between January 2007 and December 2018. Baseline characteristics and outcomes were assessed. Propensity score matching was used to compare outcomes. The primary end point was in-hospital mortality. Secondary outcomes include limb ischemia, limb ischemia interventions, distal perfusion cannula placement, stroke, bleeding, and continuous venovenous hemofiltration initiation. RESULTS: There were 394 male patients (69%) and 180 female patients (31%). After adjusting for baseline differences, propensity score matching compared 171 male patients with 171 female patients. No difference was seen between men and women in in-hospital mortality (60.2% vs 56.7%; P = .59), limb ischemia (47.4% vs 45.6%; P = .83), limb ischemia surgery (15.2% vs 12.9%; P = .64), bleeding (49.7% vs 49.1%; P = 1), continuous venovenous hemofiltration initiation (39.2% vs 32.7%; P = .26), and stroke (8.2% vs 9.4%; P = .85). Multivariable logistic regression showed that female patients who died were more likely to have had chronic kidney disease (odds ratio [OR], 2.67; 95% confidence interval [CI], 1.09-6.53; P = .032) than surviving women. Male patients who died were more likely to have had coronary artery disease (OR, 2.25; 95% CI, 1.34-3.78; P = .002) and higher lactate levels (OR, 1.14; 95% CI, 1.08-1.21; P < .001) than surviving men. Women with cardiac transplant primary graft dysfunction were more likely to survive (OR, 0.04; 95% CI, 0.01-0.27; P = .001), whereas men with cardiac transplant primary graft dysfunction were less likely to survive (OR, 0.28; 95% CI, 0.11-0.71; P = .007), than patients with other shock etiologies. CONCLUSIONS: After adjusting for baseline difference, there was no difference in outcomes between male and female patients despite differing risk profiles for in-hospital mortality.

Bleeding and Thrombotic Events During Extracorporeal Membrane Oxygenation for Postcardiotomy Shock.

BACKGROUND: Anticoagulation therapy management during venoarterial extracorporeal membrane oxygenation (ECMO) is particularly difficult in postcardiotomy shock patients given a significant bleeding risk. We sought to determine the effect of anticoagulation treatment on bleeding and thrombosis risk for postcardiotomy shock patients on ECMO. METHODS: We retrospectively reviewed patients who received ECMO for postcardiotomy shock from July 2007 through July 2019. Characteristics of patients who had bleeding and thrombosis were investigated, and risk factors were assessed by multilevel logistic regression. RESULTS: Of the 152 patients who received ECMO for postcardiotomy shock, 33 (23%) had 40 thrombotic events and 64 (45%) had 86 bleeding events. Predictors of bleeding were intraoperative packed red blood cell transfusion (odds ratio [OR] 1.05; 95% confidence interval [CI], 1.01 to 1.09), platelet transfusion (OR 1.10; 95% CI, 1.05 to 1.16), international normalized ratio (OR 1.18; 95% CI, 1.02 to 1.37), and activated partial thromboplastin time greater than 60 seconds (OR 2.32; 95% CI, 1.14 to 4.73). Predictors of thrombosis were anticoagulant use (OR 0.39; 95% CI, 0.19 to 0.79), surgical venting (OR 3.07; 95% CI, 1.29 to 7.31), hemoglobin (OR 1.38; 95% CI, 1.06 to 1.79), and central cannulation (OR 2.06; 95% CI, 1.03 to 4.11). The daily predicted probability of thrombosis was between 0.075 and 0.038 for patients who did not receive anticoagulation and decreased to between 0.030 and 0.013 for patients who received anticoagulation treatment at activated partial thromboplastin times between 25 and 80 seconds. CONCLUSIONS: Anticoagulation therapy can reduce thromboembolic events in postcardiotomy shock patients on ECMO, but bleeding risk may outweigh this benefit at activated partial thromboplastin times greater than 60 seconds.

Stroke patterns and cannulation strategy during veno-arterial extracorporeal membrane support.

Stroke has potentially devastating consequences for patients receiving veno-arterial extracorporeal membrane support (VA-ECMO). Arterial cannulation sites for VA-ECMO include the ascending aorta, axillary artery, and femoral artery. However, the influence of cannulation site on stroke risk has not been well described. The purpose of this study was to investigate the association between occurrence and patterns of stroke with ECMO arterial cannulation sites. We retrospectively reviewed 414 consecutive patients who received VA-ECMO support for cardiogenic shock between March 2007 and May 2018. Patients were categorized by cannulation strategy. The rates, subtype and location of strokes as assessed by neuroimaging during and after VA-ECMO support were analyzed. Median age was 61 years (IQR 50-69); 67% were men. 77 patients were cannulated via the ascending aorta (17%), 31 via the axillary artery (7%), and 306 (69%) via the femoral artery. In total, 26 patients (6.3%) developed 30 stroke lesions at a median of 6.0 (IQR 3.1-8.7) days after ECMO cannulation. Ischemic stroke was the most common subtype (64%), followed by hemorrhagic transformation (20%) and hemorrhagic stroke (16%). Location by CT was right hemispheric in 38%, left hemispheric in 24%, bilateral in 21%, and vertebrobasilar in 17%. The incidence of stroke was similar across cannulation strategies: aorta (n = 5, 6.5%), axillary artery (n = 2, 6.5%), and femoral artery (n = 19, 6.2%), (p = 0.99). Incidence of stroke does not appear to differ among patients cannulated via the ascending aorta, axillary artery, or femoral artery. Ischemic stroke was the most common subtype of stroke.

Impact of Venoarterial Extracorporeal Membrane Oxygenation Flow on Outcomes in Cardiogenic Shock.

Venoarterial extracorporeal membrane oxygenation (VA ECMO) is used to provide cardiopulmonary support in cardiogenic shock; however, high extracorporeal flow may increase left ventricular (LV) afterload leading to LV distention and intracardiac stasis. It is unclear how ECMO flow affects patient outcomes and complications related to ECMO. Retrospective review of patients at a single institution placed on VA ECMO from 2007 to 2018 was performed. Patients were divided into full flow (flow index > 2.2 L/min/m2) and partial flow (flow index < 2.2 L/min/m2) groups. In-hospital mortality and markers of end-organ perfusion were compared between groups balanced for risk factors using propensity score inverse probability of treatment weighting. ECMO-related complications such as LV distention, limb ischemia, and bleeding were recorded. There were 488 patients included, 405 (83%) in the partial flow group, and 83 (17%) in the full flow group. No major differences in age, gender, or comorbidities were found. There was no difference in in-hospital mortality between groups (51% vs. 55%, p = 0.59). At 72 hours post-ECMO initiation, there was no difference in the change in renal, hepatic function, or lactate from baseline nor in the rates of continuous venoveno hemofiltration initiation (p = 0.41). There was a trend towards the decreased incidence of LV distention requiring LV vent placement in the partial flow group (12% vs. 7%, p = 0.16). Compared with full flow VA ECMO, partial flow VA ECMO in carefully selected patients results in similar in-hospital mortality and provides similar end-organ perfusion for the treatment of refractory cardiogenic shock.