Compressed sensing (CS) is a rapidly developing technique that can acquire functional cines of the heart within seconds while free-breathing and it is ideal for assessing cardiac function in non-typical conditions such as when patients are sedated or anaesthetized or undergoing stress exams. CS cines can further include retrospective temporal resolution interpolation (INTPTR) to improve the frame count per heartbeat, and the impact of INTPTR on biventricular functional measurements is unknown. We investigated the impact of INTPTR on left and right ventricular volumetry and strain measurements of CS cines. Nineteen patients with 51 different CS acquisitions were prospectively enrolled. CS cines were acquired at rest, under adenosine stress, oxygen inhalation or while under general anaesthesia with mechanical ventilation. From the same CS acquisition, a dataset with and without INTPTR were generated by the scanner. The outputs were separated and analysed by blinded readers for left and right ventricular volumetry, as well as systolic and diastolic strain parameters using feature-tracking techniques. Measurements were compared between the INTPTR and non-INTPTR outputs. Similar measurements were obtained for biventricular volumes and ejection fraction independent of INTPTR. Peak strain was significantly underestimated on INTPTR cines for both longitudinal and circumferential orientations (p < 0.01). Nevertheless, good-to-excellent correlations were observed between the two measurements (r > 0.65, p < 0.01), and there was still a high area under the curve (AUC ≥ 0.95, p < 0.01) for detecting abnormal patients defined by strain analysis on the standard segmented cine. INTPTR especially negatively influenced strain rates analysis, as many strain rate curves were deemed unusable with this technique. These findings were consistent independent if the patient was in a resting, stress or anaesthetized condition. Although INTPTR is a feature which improves temporal resolution on CS cines, quantification of biventricular strain and strain rates is not feasible or comparable, thus, feature tracking analysis should be performed on non-INTPTR data. However, volumetry and ejection fraction analysis are consistent independent of which output is analysed.

BACKGROUND:Hyperglycemia is common in patients undergoing cardiovascular surgery with cardiopulmonary bypass. We hypothesize that intraoperative hyperglycemia may be, at least partially, attributable to insulin loss due to adhesion on artificial surfaces and/or degradation by hemolysis. Thus, our primary aim was to investigate the loss of insulin in 2 different isolated extracorporeal circulation circuits (ECCs), that is, a conventional ECC (cECC) with a roller pump, and a mini-ECC (MiECC) system with a centrifugal pump. The secondary aim was to assess and compare the relationship between changes in insulin concentration and the degree of hemolysis in our 2 ECC models. METHODS:Six cECC and 6 MiECC systems were primed with red packed blood cells and thawed fresh-frozen plasma (1:1). Four additional experiments were performed in cECC using only thawed fresh-frozen plasma. Human insulin (Actrapid) was added, targeting a plasma insulin concentration of 400 mU/L. Insulin concentration and hemolysis index were measured at baseline and hourly thereafter. The end points were the change in insulin level after 4 hours compared to baseline and hemolysis index after 4 hours. The insulin concentration and hemolysis index were analyzed by means of a saturated linear mixed-effect regression model with a random offset for each experiment to account for the repeated measure design of the study, resulting in mean estimates and 95% confidence intervals (CIs) of the primary end points as well as of pairwise contrasts with respect to ECC type. RESULTS:Insulin concentration decreased by 63% (95% CI, 48%-77%) in the MiECC and 92% (95% CI, 77%-106%) in the cECC system that contained red blood cells. Insulin loss was significantly higher in the cECC system compared to the MiECC ( P = .022). In the cECC with only plasma, insulin did not significantly decrease (-4%; 95% CI, -21% to 14%). Hemolysis index in MiECC increased from 68 (95% CI, 46-91) to 76 (95% CI, 54-98) after 4 hours, in cECC from 81 (95% CI, 59-103) to 121 (95% CI, 99-143). Hemolysis index and percent change of insulin showed an excellent relationship (r = -0.99, P < .01). CONCLUSIONS:Our data showed that insulin levels substantially decreased during 4 hours of simulated cardiopulmonary bypass only in the ECC that contained hemoglobin. The decrease was more pronounced in the cECC, which also exhibited a greater degree of hemolysis. Our results suggest that insulin degradation by hemolysis products may be a stronger contributor to insulin loss than adhesion of insulin molecules to circuit surfaces.

Objectives To examine whether estimates of peak global systolic (S ') and diastolic (E ') left ventricular (LV) flow rates based on 3D echocardiographic volumes are feasible and match physiology.Methods In this retrospective feasibility study, we included patients undergoing major cardiac surgery. S ' and E ' were derived from 190 patients by taking the first derivative of the volume-time relationship of 3D ecg-gated transesophageal echocardiography (TEE) images. To examine the quality of images upon which the estimates of flow were based we correlated intraoperative 3D TEE and preoperative 2D transthoracic echocardiography (TTE) volumes. As a proof-of-concept, we then correlated S ' flow with stroke volume and S ' and E ' were compared by valve pathology.Results In each of the 190 images, S ' and E ' were derived. There was good correlation between 1) the ejection fraction (EF) of 3D LV images obtained intraoperatively by TEE and preoperatively by TTE (Pearson's r = 0.65) and also 2) S ' and stroke volume (Pearson's r = 0.73). Patients with aortic or mitral regurgitation showed higher S ' than patients without valve pathologies (-315 mL/s [95% CI -388 mL/s to -264 mL/s]P = 0.001, -319 mL/s [95% CI -397 mL/s to -246 mL/s]P = 0.001 vs -242 mL/s [95% CI -300 mL/s to -196 mL/s]). These patients also showed higher E ' than patients without valve pathologies (302 mL/s [95% CI 237 mL/s to 384 mL/s]P = 0.006, 341 mL/s [95%CI 227 mL/s to 442 mL/s]P = 0.001 vs 240 mL/s [95%CI 185 mL/s to 315 mL/s]). Patients with aortic stenosis showed no difference in S ' or E' (-263 mL/s [95%CI -300 mL/s to -212 mL/s]P = 0.793, 255 mL/s [95%CI 188 mL/s to 344 mL/s]P = 0.400).Conclusions Estimates of global peak systolic and diastolic LV flow based on 3D TEE are feasible, promising, and match valve pathologies.

Objective Mechanical invasive intermittent positive pressure ventilation (IPPV) has known cardiovascular effects that differ across the heart and in particular poses challenges in patients with right heart dysfunction. Invasive measurements have traditionally been used to study its haemodynamic impact. Yet these parameters have often been acquired sequentially and don't provide a comprehensive picture. Whole heart haemodynamics can now be quantified non-invasively using cardiovascular magnetic resonance 4D-flow imaging. From a single acquisition, many parameters are quantified including blood flow turbulence, volumes, velocities, kinetic energies, and pressure gradients in any orientation. To gain insight into the effects of IPPV on cardiovascular haemodynamics, we applied 4D-flow imaging in patients before and during general anaesthesia with IPPV. Design and method Six patients (ASA I-III) scheduled for elective surgery without preoperative heart failure, valve, or significant respiratory diseases were prospectively recruited. In a magnetic resonance scanner, t4D-flow imaging datasets were taken: 1) while patients were spontaneously breathing just prior to induction, 2) fifteen minutes after the induction of general anaesthesia and installation of IPPV, prior to any surgical procedure. IPPV was set to tidal volumes of 6ml/kg and ventilation rate adjusted for normocapnic petCO2 35-40mmHg. Results and conclusions Blood pressure decreased from 133±7/76±4mmHg at baseline to 96±2/54±2mmHg (p<0.01) at IPPV acquisition. As a marker of preload, end-diastolic volume dropped in the LV (148±12 to 125±11ml, p<0.01), while the RV drop was non-significant (176±18 to 165±20ml, p=0.08). Cardiac output (CO) was subsequently reduced (4.5±0.4 to 3.4±0.3L/min, p=0.03). As markers of afterload, aortic kinetic energy non-significantly decreased (216±31 to 163±77µJ/ms, p=0.06). Kinetic energy decreases through reduced CO and higher resistance, and a deeper analysis showed that aortic resistance (28±5 to 19±3dynescm-5, p=0.03) and maximum systolic pressure gradient across the aortic valve significantly dropped (7.1±1.0 to 4.0±0.7mmHg, p<0.01) allowing for a minimal kinetic energy change despite lower flow (350±27 to 282±24ml/s, p=0.02). For the right heart, kinetic energy of the pulmonary artery significantly dropped (182±15 to 107±9µJ/ms, p=0.03). Unlike the aorta, systolic pressure gradients across the pulmonary valve (3.2±0.1 to 3.0±0.5mmHg, p=0.68) and resistance (13±1 to 16±3dynescm-5, p=0.29) didn't drop despite reduced flow (335±16 to 261±31ml/s, p=0.01). In fact, the 26±21% increase in resistance of the pulmonary artery was significantly higher than the -31±7% drop observed in aorta (p=0.01). This corresponded to changes in ventricular systolic kinetic energy. While both ventricles had a drop in kinetic energy with IPPV likely due to reduced CO (LV: 33±4 to 24±2uJ/ml p=0.05, RV: 37±4 to 16±2uJ/ml, p<0.01), the percent-change of RV kinetic energy loss was greater than the LV (-66±6% vs -25±8%, p=0.04), reflecting the difference in afterload changes between the left and right heart.In conclusion, 4D-flow magnetic resonance imaging non-invasively quantifies the effects of anaesthesia induction and IPPV on haemodynamics across the heart using known and unique parameters such as kinetic energy. In particular, IPPV causes different afterload and intraventricular haemodynamics changes in the right heart. Research applications of 4D-flow in a perioperative environment may help the field of anaesthesia understand how the heart responds to IPPV and how potential cardiovascular complications could be mitigated.

Objective Many anaesthetic agents used during the induction and maintenance of general anaesthesia (GA) have known cardiac depressant properties. Anaesthetists try to find a balance between sufficient anaesthesia depth and minimizing cardiovascular side effects. Especially in patients with underlying heart disease, this balance is more delicate to maintain. Multiple imaging modalities such as echocardiography and cardiovascular magnetic resonance (CMR) provide a non-invasive assessment of ventricular function. Strain analysis of these images can provide sensitive measurements of contractility through peak strain. Another unexplored marker from strain analysis is ventricular mechanical dyssynchrony (MD). MD is the non-uniformity in the timing of contraction between the myocardial segments, and it may provide insight into adverse perioperative cardiac dysfunction as dyssynchrony is known to attenuate ventricular ejection efficiency. In a research setting, CMR can acquire high quality images of the ventricle continuously without requiring breath-holds. This study implemented CMR imaging in a perioperative environment to quantify time-resolution changes in ventricular contractility and synchronicity in patients undergoing GA. Design and method Eight patients (ASA I: n=3, ASA II: n=3, ASA III, n=2) scheduled for elective operations under GA, were prospectively recruited to undergo a CMR exam during anaesthesia induction. Propofol (2mg/kg) was used for induction while anaesthesia was maintained with sevoflurane. The heart was imaged using real-time cine loops of the left ventricle in a short-axis slice at a baseline state when patients were awake prior to preoxygenation, and continuously from intubation until 10 minutes after intubation. On these images, global circumferential strain and MD (standard deviation of the time to peak strain of the myocardial segments) of the left ventricle were assessed. Values are reported as median [interquartile range]. Results and conclusions Neither peak strain (-18.6% [-20.0, -17.2] to -18.0% [-18.3, -17.0], p=0.06) nor MD (20ms [6, 25] to 14ms [5, 24], p=0.42) changed significantly from baseline to images obtained immediately after intubation. However, after intubation strain consistently worsened (-15.3% [-17.6, -14.4]) and by 10 minutes of anaesthesia maintenance was significantly poorer than both baseline (p<0.01) and immediately after intubation (p=0.03). Similarly, MD significantly rose, indicating ventricular dyssynchrony (25ms [19, 41], p=0.03 vs baseline, p=0.04 vs post-intubation). As can be seen in the figure, the two patients with cardiovascular disease (ASA III) had the highest change in MD over the first 10min of maintenance, while no such distinction in peak strain is observed between ASA classes in this interim sample.Using continuous CMR imaging techniques, a gradual and significant decline in ventricular function was observed in the first ten minutes after GA induction. Not only did peak strain decline, but simultaneous assessment of MD demonstrated left ventricular contraction became more dyssynchronous as well. These parameters can both be acquired by multiple imaging modalities, including perioperative transoesophageal echocardiography. Ventricular mechanical dyssynchrony may provide further insight into understanding the effect of GA on cardiac function, and how this may lead to the onset of cardiac complications in high-risk patients.

IntroductionThe cardiovascular system is vulnerable to effects associated with induction of general anaesthesia (GA). Novel free-breathing cardiovascular magnetic resonance (CMR) techniques can now image rapid fluctuations in tissue oxygenation and wall function of the myocardium beat-by-beat. This provides an excellent temporal resolution to spatially map myocardial oxygenation changes. We present interim findings from an innovative study applying oxygenation-sensitive cardiovascular magnetic resonance (OS-CMR) imaging to investigate the dynamics of myocardial oxygenation and function during the induction of GA in patients with and without coronary artery disease (CAD).MethodsFour patients without cardiovascular risk factors (no-CAD) scheduled for elective orthopaedic surgery (ASA-PS class I or II), and one patient with documented single-vessel CAD scheduled for elective minimally invasive direct coronary artery bypass surgery (ASA-PS class III) underwent GA induction inside an MRI scanner. OS-CMR imaging was applied continuously from the awake state throughout the entire induction process and for 10 minutes of anaesthesia maintenance, after which patients were transferred to their scheduled surgery. Changes in myocardial oxygenation were calculated in comparison to the awake state (%-change from baseline OS-CMR signal intensity). From these same images, myocardial function was quantified in a circumferential orientation using feature tracking strain analysis.ResultsFirst with pre-oxygenation (14 breaths/min paced breathing / FiO2=1.0), petCO2 (No-CAD:27±1mmHg, CAD:26mmHg) and myocardial oxygenation (No-CAD:-8.6±1.6% CAD:-8.8%) decreased in all patients. However, while the No-CAD patients showed an increase in myocardial contractility (strain:-16.3±1.1% to -17.5±11.7%), strain worsened during this phase in the CAD patient (-19.1% to -15.8%). With administration of sufentanil, breathing slowed with subsequent normalisation of petCO2 (No-CAD:37±2mmHg, CAD:35mmHg), myocardial oxygenation (no-CAD:2.4±4.1%, CAD:0.5%) and a partial recovery in strain (No-CAD:-15.1±0.3%, CAD:-16.8%). After propofol and rocuronium administration, manual bag ventilation led to another decrease in petCO2 (no-CAD:23±3mmHg, CAD:20mmHg), a decrease in myocardial oxygenation (No-CAD:-5.2±3.0, CAD:-11.0%), no change in strain in heart-healthy patients (-16.4±0.7%) but a decline in ventricular function in the CAD patient (strain=-15.0%). During maintenance (ventilation rate adjusted to target petCO2 35-40mmHg / FiO2=0.4) longer phases of myocardial de-oxygenation (OS-CMR<0%) were observed until 8.7±0.8min after intubation in the no-CAD patients, and until 8.5min in the CAD patient, respectively. By 10min after intubation, myocardial oxygenation had normalized to baseline levels but strain remained reduced (No-CAD:-14.0±10.5, CAD:-11.7%), likely due to direct cardio-depressant effects of sevoflurane. In the CAD patient, co-localized myocardial oxygenation and strain abnormalities were especially pronounced in the myocardial subtended to a haemodynamically significant stenosis, while No-CAD patients exhibited homogenous changes in the myocardium.DiscussionInterim results demonstrate that in the first study to ever use OS-CMR to monitor the heart during GA induction, complex fluctuations of myocardial oxygenation and function occur in patients with and without CAD. The patient with single-vessel CAD exhibited regional deoxygenation co-localized with myocardial strain deterioration in the territory at risk, indicative of inducible regional myocardial ischaemia. Implementation of advanced imaging in an anaesthesia research setting allows us to investigate how the heart responds to potential perioperative triggers of inducible ischaemia and may further advance the practice of anaesthesia for cardiac risk populations.

Background High-flow nasal oxygenation is increasingly used during sedation procedures and general anesthesia in apneic patients. Transcutaneous CO 2 (p tc CO 2 )-monitoring is used to monitor hypercapnia. This study investigated p tc CO 2 -monitoring during apneic oxygenation. Methods We included 100 patients scheduled for elective surgery under general anesthesia in this secondary analysis of a randomized controlled trial. Before surgery, we collected p tc CO 2 measured by TCM4 and TCM5 monitors and arterial blood gas (ABG) measurements every two minutes during 15 minutes of apnea. Bland-Altman plots analyzed agreement between measurement slopes; linear mixed models estimated the different measuring method effect, and outlined differences in slope and offset between transcutaneous and arterial CO 2 partial pressures. Results Bland-Altman plots showed a bias in slope (95% confidence intervals) between ABG and TCM4-measurements of 0.05mmHg/min (-0.05 to 0.15), and limits of agreement were -0.88mmHg/min (-1.06 to -0.70) and 0.98mmHg/min (0.81 to 1.16). Bias between ABG and TCM5 was -0.14mmHg/min (-0.23 to -0.04), and limits of agreement were -0.98mmHg/min (-1.14 to -0.83) and 0.71mmHg/min (0.55 to 0.87). A linear mixed model (predicting the CO 2 -values) showed an offset between arterial and transcutaneous measurements of TCM4 (-15.2mmHg, 95%CI: -16.3 to -14.2) and TCM5 (-19.1mmHg, -20.1 to -18.0). Differences between the two transcutaneous measurements were statistically significant. Conclusions Substantial differences were found between the two transcutaneous measurement systems, and between them and ABG. Transcutaneous CO 2 monitoring cannot replace arterial CO 2 -monitoring during apneic oxygenation. In clinical settings with rapidly changing CO 2 -values, arterial blood gas measurements are needed to reliably assess the CO 2 -partial pressure in blood. Trial registration ClinicalTrials.gov ( NCT03478774 ).

Editor—In patients with cardiac disease, induction of general anaesthesia is a vulnerable phase, and it bears many potential triggers for provoking myocardial ischaemia. It involves changes in respiration, medications, tracheal intubation and initiation of positive-pressure ventilation, all of which can lead to rapid fluctuations in haemodynamics and arterial blood gases, with potential shifts in the myocardial oxygen supply–demand balance. 1 Priebe H.-J. Triggers of perioperative myocardial ischaemia and infarction. Br J Anaesth. 2004; 93: 9-20 Abstract Full Text Full Text PDF PubMed Scopus (121) Google Scholar Perioperatively, the majority of myocardial injury can be attributed to oxygen-supply mismatch, 2 Puelacher C. Gualandro D.M. Glarner N. et al. Long-term outcomes of perioperative myocardial infarction/injury after non-cardiac surgery. Eur Heart J. 2023; : ehac798 Google Scholar which is often clinically silent yet one of the first stages of the ischaemic cascade. 3 Rafiudeen R. Barlis P. White H.D. van Gaal W. Type 2 MI and myocardial injury in the era of high-sensitivity troponin. Eur Cardiol Rev. 2022; 17: e03 Crossref PubMed Scopus (4) Google Scholar To our knowledge, the time course of myocardial oxygenation during induction of general anaesthesia has not been studied.