Abstract Introduction Acute right ventricular (RV) failure is common in critically ill patients and is associated with poor outcomes. Mechanical ventilation (MV) in patients with RV failure is challenging and lacks supporting evidence. Permissive hypercapnia may offer physiological benefits, allowing a reduction in intrathoracic pressures, thereby mitigating the impact of MV on RV function. Although hypercapnia is associated with increased pulmonary vascular resistance (PVR), it remains unclear whether this results from hypercapnia itself or the accompanying respiratory acidosis. We hypothesised that permissive hypercapnia may be beneficial in RV failure; buffering acidosis enabled assessment of the separate effects of hypercapnia and acidaemia. Methods Prospective, randomized crossover preclinical study in ovine models of acute severe RV failure induced by either 1) pulmonary artery banding to rapidly increase RV afterload, targeting an RV systolic pressure >50 mmHg or 2) myocardial injury through direct injections of 96% ethanol into the RV free wall, targeting an RV ejection fraction <35%. Sheep were ventilated for 30 minutes with different settings as reported in Table 1. Briefly, following baseline assessment, permissive hypercapnia and buffered permissive hypercapnia were sequentially applied. A Swan-Ganz catheter was used to monitor systemic and pulmonary haemodynamics. Mechanical Power (MP) was computed according to previous formula1. The primary outcome was RV ejection fraction computed through a pressure-volume (PV) loop. Secondary outcomes were biventricular PV loop; cardiac and lung ultrasound; respiratory mechanics; lung, heart and kidney histology. Results We report preliminary results from one experiment in which a sheep underwent pulmonary artery banding. Across the baseline, permissive hypercapnia, and buffered permissive hypercapnia timepoints, mechanical power was 29, 11, and 14 J/min, respectively. The right ventricular ejection fraction was 37%, 35%, and 42%, while pulmonary vascular resistance was 8.8, 8.2, and 6.3 Wood units. Cardiac output measured 3.5, 4.4, and 4.4 L/min, with corresponding stroke volumes of 36, 37, and 45 mL. Conclusions To the best of our knowledge, this is the first assessment of buffered permissive hypercapnia in a model of RV dysfunction. Preliminary findings indicate that buffering respiratory acidosis improves RV performance and permits ventilation at low mechanical power. 1. Gattinoni L. et al, Intensive Care Med. 2016;42(10):1567-75 This abstract is funded by: The Prince Charles Hospital Foundation

Long-term sequelae associated with COVID-19 are currently poorly defined, specifically in patients who have been discharged from intensive care units. We hypothesized that functional and pulmonary capacity, as well as mental health and cognitive function, would remain outside the normal range for up to 12 months after ICU discharge. Methods This is an international multicenter, prospective follow-up study of patients admitted to ICU due to COVID-19 (1). At 3, 6 and 12 months after ICU discharge, the following assessments were done: 6-minute walking test, Barthel Index Score (BIS), WHODAS score, spirometry, lung diffusion capacity for carbon monoxide (DLCO), Hospital Anxiety and Depression Scale (HADS), Patient Health Questionnaire-9 (PHQ-9) and the Montreal Cognitive Assessment (MoCA). Normalized test results at any follow-up assessment were not repeated at subsequent visits. Results 354 patients from 9 sites in 6 countries (Ireland, Spain, Italy, USA, Singapore, Colombia) admitted to ICU from January 2021 to December 2022 for COVID-19 were included. Median (IQR) age was 60 years (SD 13.3y), 32.5% were females, and 22 patients (6.2%) died during follow-up. The 6-min walking test was outside a normal range in 97.7%, 98% and 99.1% of the patients at 3, 6 and 12 months, with a mean walking distance of 359, 392 and 433 meters, respectively (Fig.1A). The BIS score was not normalized in 12.4%, 23.5% and 28% (Fig.1B), the WHODAS score in 22.2%, 56.4% and 45.2% (Fig.1C) at 3, 6 and 12 months, respectively. FVC and FEV1 were outside normal limits in 64.3% of the patients at 3 months, 34.7% and 32.9% at 6 months, 33.8% and 27.7% at 12 months (Fig.1D/E). DLCO was abnormal in 62.4%, 78.4% and 80.6% (Fig.1F) of patients, at 3, 6 and 12 months. As shown by the HADS, anxiety was present in 28.9%, 26.9% and 32.7% (Fig.1G) of the patients, while 30.2%, 26.9% and 32.7% (Fig.1H) had significant depression at 3, 6 and 12 months. The PHQ-9 was outside normal range in 66.4%, 66.2% and 72.2% (Fig 1I) of the patients, while the MOCA test revealed cognitive impairment in 40.8%, 30% and 32.5% of the patients at 3, 6 and 12 months. Conclusion The assessment of COVID-19 patients up to 12 months post-ICU discharge indicates that pulmonary function and mental health remain outside the normal range in a significant proportion of patients, with one-third of patients experiencing cognitive impairment. A comprehensive evaluation of the pre-ICU admission status is currently underway.

RATIONALE Since the initial clinical characterization of Acute Respiratory Distress Syndrome (ARDS), extensive research efforts utilizing animal models of acute lung injury (ALI) have aimed to elucidate its pathophysiology and evaluate potential treatments. However, findings from these animal models have often failed translation into clinical practice, highlighting the need for refined research methodologies. In 2011, the American Thoracic Society (ATS) published a landmark workshop report, defining ALI in experimental models based on structural, functional, biochemical, and histopathological domains. The report was updated in 2021, recommending that mechanistic studies should address at least one of these four domains, while preclinical modelling studies with greater translational intent should address at least three domains. Finally, studies testing therapeutic interventions are meant to address all four. This scoping review focuses on large animal models of ALI, which are considered more suitable for clinical translatability, investigating their adherence to 2021 ATS guidelines. METHODS Following PRISMA guidelines, we searched three indexed databases (PubMed/MEDLINE, EMBASE, CINAHL) for studies published or translated to English between January 2011 and August 2022. The search strategy included terms related to large-animal models and ALI/ARDS. Abstracts were screened for relevance, and full-text manuscripts independently reviewed. Based on the 2021 ATS framework, we categorized studies as either mechanistic or preclinical modelling. Preclinical modelling studies were further divided into new ALI model development, drug testing, and intervention testing. RESULTS We identified 4,837 studies, of which 507 met the inclusion criteria. Detailed information on animal types, number of animals, study groups, animal weight and age, and types of pulmonary injuries are presented in Table 1a. Among the included studies, 204 (40.2%) focused on the appraisal of a specific aspect of ARDS pathophysiology (mechanistic studies), while 303 (59.7%) were categorized as preclinical modelling, including new ALI model development (12.8%), drug testing (23.6%), and intervention testing (23.3%). Mechanistic studies showed high adherence to ATS guidelines, with 93.1% addressing at least one domain (Table 1b). Conversely, only 36.9% of ALI model development studies complied with the recommendation of addressing three domains. Adherence was even lower for studies testing drugs (33.3%) and interventions (26.2%), which should have addressed all four domains. CONCLUSIONS This review highlights poor adherence to 2021 ATS guidelines across peer-reviewed large-animal studies of ALI, specifically in studies evaluating novel drugs and interventions. We call for an urgent improvement in study methods for ALI to enhance the translatability of preclinical findings and ultimately improve patient outcomes.

Rationale: Lung protective strategies in managing Acute Respiratory Distress Syndrome (ARDS) often comprise interventions to recruit and keep diseased lung regions open, thereby improving oxygenation and avoiding Ventilator Induced Lung Injury (VILI). CT scans and Electrical Impedance Tomography are often used to assess lung recruitment and de-recruitment. However, there is a need for non-invasive, precise tools to predict lung recruitability. This study explores the novel application of heartbeat interference signals, a novel parameter recorded from ventilator waveforms referring to the heart beating against the surrounding lung, as a potential non-invasive indicator of lung response to recruitment manoeuvres and different Positive End Expiratory Pressure (PEEP) levels in ARDS. Methods: We retrospectively analysed data from the STARDUST preclinical trial on an ARDS ovine model – performed at the Critical Care Research Group, Brisbane, Australia. Quantitative data on lung recruitment and overdistension were obtained from CT-scans performed at two different PEEP levels (5-45 cmH2O), after 48 hours of mechanical ventilation (MV). We defined recruitability potential as the change in aerated lung volume between PEEP of 45 and 5 cmH2O, normalised to the aerated volume at PEEP 5cmH2O. We classified subjects as ‘recruiters’ and ‘over-distenders’ if their recruitability and overdistension potential, respectively, exceeded the mean value for the analyzed population. Heartbeat interference parameters were extracted from the ventilator waveforms before the CT-scan. We defined the Heartbeat Inteference Amplitude during inspiration and expiration (HBI-AInsp, HBI-AExp) as the maximum amplitude of the waveform generated by each heartbeat on the ventilator pressure waveform during inspiratory and expiratory pauses. A comparison in recruiters vs non-recruiters, and overdistenders vs non-overdistenders was made on 5 sheep with Mann-Whitney U test. Results: Overall, the CT scans confirmed severe ARDS in the studied population, with poorly aerated and non-aerated tissue from 68% at PEEP 5 cmH2O to 41% at PEEP 45 cmH2O. The analysis of the HBI parameters revealed no statistically significant differences (p-value = 0.2) in HBI-AInsp between recruiters and non-recruiters, although a trend toward higher values in recruiters was observed. No other parameters were associated with lung recruitability and overdistension. Conclusions: Our findings suggest a possible role for HBI parameters in discriminating between recruiters and non-recruiters. Further analysis involving data from the ventilator during the CT-scan execution and, even more importantly, lung mechanics and systemic oxygenation outcomes after the increase in PEEP will further expand our understanding of these novel indexes.

RATIONALE The initial lung response to injury is characterized by the activation of innate immune cells, leading to damage of the alveolar endothelial and epithelial barriers and flooding of the alveolar space with inflammatory biofluids. While the increased production and displacement of these biofluids across the airways in early acute lung injury (ALI) are well documented, their potential role in propagating inflammation to non-injured lung regions remains largely unexplored. This study aims to investigate how biofluids may contribute to the propagation of lung injury through the airway dissemination of inflammatory mediators, introducing a novel pathogenic mechanism termed Biofluid-Induced Lung Injury (BILI). METHODS Six female donor pigs underwent bronchoalveolar lavage (BAL) fluid collection before and after oleic acid-induced ALI to obtain non-inflammatory and inflammatory biofluids, respectively. The collected biofluids were centrifuged to separate cellular components, and the supernatant was stored at -80°C. In the second experimental phase, six healthy female recipient pigs were instilled with biofluids from a single randomly assigned donor. Recipients were positioned prone, and biofluids were instilled into the left upper lobe and right middle lobe, selected for their similar spatial orientation. Each recipient consistently received non-inflammatory biofluids in one lobe and inflammatory biofluids in the other, with the lobe assignment randomized across animals. Recipients underwent nine instillations of 20mL aliquots of biofluids—one every three hours, up to 24 hours. Three hours after the final instillation, BAL was performed on each of the five lobes of the recipients. Cytokine concentrations in both donor and recipient BAL fluid were quantified using Luminex Magpix-based assays (R&D System). RESULTS Donor pigs (58.8 ± 3.4kg) successfully developed ALI after receiving 0.08 ± 0.015mL/kg of intravenous oleic acid. Of six measured cytokines (TNF-α, IL-10, IFN-γ, IL-1β, IL-8, IL-6), none were detectable in the non-inflammatory BAL fluid, while IL-8 and IL-6 were present in the inflammatory one, with median concentrations of 89.75pg/mL (IQR 17-111pg/mL) and 354.3pg/mL (IQR 306-622pg/mL), respectively. Recipient pigs (61.7 ± 3.9kg) showed no detectable cytokines in non-instilled lobes. A trend toward increased cytokine concentrations in BAL fluid from lobes instilled with inflammatory biofluids was found, particularly for IL-6, compared with BAL fluid from lobes instilled with control biofluids (Table 1). CONCLUSIONS Preliminary data suggests that inflammatory biofluids from early ALI may incite new inflammatory processes in non-injured lung regions, supporting the injurious capability of BILI. Further studies, including histopathological analysis, are underway to confirm this evidence.

Rationale: Acute Respiratory Distress Syndrome (ARDS) is characterised by inhomogeneous lung damage, with the more compromised alveolar units leading to increased strain over the healthy ones, exacerbating Ventilator-Induced Lung Injury (VILI). Only a few monitoring tools enable clinicians to capture this heterogeneity, albeit suboptimally. Respiratory system compliance (Crs), for example, is a surrogate of whole-lung injury severity, while tools like Electrical Impedance Tomography (EIT) and CT scanning provide more precise regional insights but are time-consuming and challenging to interpret. Without a precise picture of regional differences, individualised mechanical ventilation remains utopic. Therefore, there is a need for non-invasive, bedside available, easily interpretable tools to predict regional lung mechanics. This study explores the novel application of heartbeat interference, the signal recorded from ventilator waveforms referring to the heart beating against the surrounding lung, as a potential non-invasive indicator of regional aeration in ARDS. Methods: We retrospectively analysed data from two preclinical trials – STARDUST and HYDROGEN-ECMO, performed at the Critical Care Research Group, Brisbane, Australia – in ARDS ovine models. 48 hours of ventilator pressure-time waveforms and ECG signals were synchronized to isolate the heartbeat interference. Bandpass filtering targeting specific frequencies (1-2 Hz) allowed for parameter extraction from the ventilator waves. These include Amplitude (HBI-Amplitude) and Heartbeat Interference Recoil (HBI-Recoil), recorded once every hour during inspiratory and expiratory pauses (see Figure for a thorough definition of the parameters). HBI-Amplitude and Recoil were correlated with traditional lung mechanics indicators (e.g. Crs, Extravascular Lung Water Index [EVLWI]) and cardiovascular indexes (e.g. Cardiac Index [CI], Global End-Diastolic Volume Index [GEDVI]) also recorded once every hour throughout the experiments. 378 measurements from 14 sheep were compared using a mixed model. Results: HBI-Amplitude during inspiration (HBI-AInsp) showed a statistically significant positive correlation with Crs (p=0.019) and a non-significant negative trend with EVLWI (p=0.364). HBI-AInsp was positively correlated with cardiac index (CI) (p=0.001) but independent of GEDVI. On the contrary, HBI-Recoil did not significantly correlate with respiratory or hemodynamic parameters. Conclusions: Our preliminary results indicate that HBI-AInsp correlates with Crs, potentially suggesting reduced heartbeat interference across stiffer lungs. Further evaluations on prospective trials with intended data extraction are crucial to implement the current findings and allow translation to clinical data interpretation. Moreover, benchmarking the current findings with regional data from EIT and CT-scan is critical to appraise the diagnostic potentials of our newly proposed parameters for bedside non-invasive assessment of regional lung aeration.

Transfusion-related acute lung injury (TRALI) is a hazardous transfusion complication with an associated mortality of 5% to 15%. We previously showed that stored (5 days) but not fresh platelets (1 day) cause TRALI via ceramide-mediated endothelial barrier dysfunction. As biological ceramides are hydrophobic, extracellular vesicles (EVs) may be required to shuttle these sphingolipids from platelets to endothelial cells. Adding to complexity, EV formation in turn requires ceramide. We hypothesized that ceramide-dependent EV formation from stored platelets and EV-dependent sphingolipid shuttling induces TRALI. EVs formed during storage of murine platelets were enumerated, characterized for sphingolipids, and applied in a murine TRALI model in vivo and for endothelial barrier assessment in vitro. Five-day EVs were more abundant, had higher long-chain ceramide (C16:0, C18:0, C20:0), and lower sphingosine-1-phosphate (S1P) content than 1-day EVs. Transfusion of 5-day, but not 1-day, EVs induced characteristic signs of lung injury in vivo and endothelial barrier disruption in vitro. Inhibition or supplementation of ceramide-forming sphingomyelinase reduced or enhanced the formation of EVs, respectively, but did not alter the injuriousness per individual EV. Barrier failure was attenuated when EVs were abundant in or supplemented with S1P. Stored human platelet 4-day EVs were more numerous compared with 2-day EVs, contained more long-chain ceramide and less S1P, and caused more endothelial cell barrier leak. Hence, platelet-derived EVs become more numerous and more injurious (more long-chain ceramide, less S1P) during storage. Blockade of sphingomyelinase, EV elimination, or supplementation of S1P during platelet storage may present promising strategies for TRALI prevention.

Ischaemia-reperfusion injury (IRI) results in myocardial cell death and compromised cardiac function making it a leading risk factor for heart failure.

Rationale: Patients with COVID-19 commonly develop severe hypoxemic respiratory failure and require invasive mechanical ventilation (MV). The disease burden and predictors of mortality in this population remain uncertain. Methods: Prospective observational cohort study from 139 intensive care units of the international COVID-19 Critical Care Consortium. Patients enrolled from January 14th through November 31st 2020 were included in the analysis. Patient's characteristics and clinical data were assessed. Multivariable Cox proportional hazards analysis was conducted to identify indipendent predictors of mortality within 28 days from commencement of MV. Results: 1578 patients on MV were included into the analysis. Mean±SD age was 59 years±13 and patients were predominantly males (66%). 542 Patients (34.4%) died within 28 days from commencement of MV. Nonsurvivors were slightly older (mean age±SD 62±13 vs. 59±13) and presented more frequently hypertension, chronic cardiac disease and diabetes. Median (IQR) PaO2/FiO2 upon commencement of MV was 96 (68-135) and 111 (81-173) in patients who did not survive vs. survivors, respectively (p=0.04). ECMO (13% vs 25%, p<0.01), inhaled nitric oxide (11% vs 15%, p=0.02) and recruitment manoeauvres (26% vs 31%, p<0.01) were used less frequently in patients who did not survive. Independent risk factors associated with 28-day mortality included age older than 70 years (hazard ratio [HR], 2.83;95% CI, 1.32-6.07), higher creatinine levels upon ICU admission (HR, 1.20;95% CI, 1.03-1.40), and lower pH within 24h from commencement of MV (HR, 0.12;95% CI, 0.02-0.62), while a shorter period (day) from early symptoms to hospitalisation reduced mortality risks (HR, 0.96;95% CI, 0.93-0.99). Conclusions: Our findings from a large international cohort of critically-ill COVID-19 patients on mechanical ventilation emphasises that elderly patients, not promptly admitted to the hospital, and who present higher creatinine levels and acidosis are at higher risk of mortality.

Purpose Cold static storage (CSS) is the standard method for heart preservation during transplantation (HTx). However, CSS beyond 4 hours increases the risk of primary graft dysfunction (PGD). Hypothermic ex vivo perfusion (HEVP) of donor hearts allows oxygen delivery during preservation, and may facilitate extended donor preservation without increasing PGD risk. We compared post-HTx survival, systemic inflammation and cardiac function following donor heart preservation by CSS (2 hrs) versus HEVP (2 and 8 hrs). Methods Brain death was induced in donor sheep for 24 hrs. Donor hearts were preserved by a) CSS for 2 hrs (n=7), b) HEVP for 2 hrs (n=4), or c) HEVP for 8 hrs (n=4). Orthotopic HTx was performed in matched recipients. Recipients were weaned from cardiopulmonary bypass and monitored for 6 hrs. Recipient blood was collected and assayed for inflammatory cytokines and cardiac markers. Cardiac function was assessed by echocardiography. Results Six-hour survival was 71% following CSS, and 100% following 2 and 8 hrs HEVP, respectively. Recipients systemic interleukin-6 and 8 levels were reduced using HEVP vs CSS. Post-HTx haemodynamic function was no different between groups, but HEVP reduced the requirement for vasoactive support compared to CSS (2 hrs CSS: 1.57±0.7; 2 hrs HEVP: 0.35±0.09; 8 hrs HEVP: 0.35±0.05 mmHg−1). HEVP was associated with reduced post-HTx lactate (2 hrs CSS: 11.4±1.8; 2 hrs HEVP: 5.2±0.7; 8 hrs HEVP: 6.7±1.3 mmol/L), more stable base excess and physiological pH in blood. Post-HTx cardiac function was no different between groups. Cardiac troponin I levels were comparable between CSS vs. 8 hrs HEVP, but reduced with 2 hrs HEVP. Conclusion Preliminary data on donor heart preservation by HEVP shows promising outcomes in comparison to CSS. Heart preservation by HEVP can be extended up to 8 hours, without compromising post-HTx recipient survival. HEVP may assist in overcoming limitations in preservation time associated with HTx, without increasing PGD risk.

RATIONALE. Neuromuscular blocking agents (NMBA) are used in patients with moderate to severe acute respiratory distress syndrome. NMBA have also been used in COVID-19 patients who required mechanical ventilation (MV), but their benefit-to-risk ratio remains uncertain.METHODS. We investigated the effects associated with the use of NMBA in COVID-19 patients who required MV from January 1, 2020 to October 31, 2020 in 153 hospitals across 6 continents, comprising the COVID-19 Critical Care Consortium. Cox proportional hazards analysis was conducted to study the impact of NMBA on 28-day intensive care unit (ICU) mortality. Hospital/ICU lengths of stay were appraised. We performed a propensity score (PS) matching analysis to control confounding factors.RESULTS. 1227 patients were eligible for analysis, among those 598 (48.7%) received NMBA for 2 days or longer, with a median time from ICU admission to commencement of NMBA therapy of 0 day (IQR 0-1 days). The median duration of NMBA therapy was 2 days (N=789, IQR 1-5). In comparison with standard of care, treatment with NMBA was more frequent in obese (31% vs. 39%, P = 0.03) and diabetic patients (2% vs. 8%, P <0.01) and less frequent in patients with hypertension (52% vs. 46%, P =0.04) or cardiac diseases (21% vs. 14%, P =0.003). Upon commencement of MV, patients who underwent NMBA therapy vs those who did not presented a PaO2/FiO2 of 136.1±69.2 vs. 162.7 ±125.8 (p<0.01), required more often ECMO (10% vs 5.2%, p <0.01) and prone position (25.1% vs 6.2%, p <0.01). Unadjusted 28-day all-cause mortality was similar (58.2% vs. 62.4%, P =0.134) between patients without or with NMBA therapy, respectively, but length of MV (3 days [2-5] vs. 6 [3-12] P <0.01) and ICU stay (8 days [4-14] vs. 13 [7-19] P <0.01) were prolonged. After PS matching, NMBA therapy was strongly associated with 28-day ICU mortality (adjusted HR 3.18, 95% CI 2.65-3.81, P <0.01). CONCLUSION. Use of NMBA in COVID-19 patients requiring MV is associated with increased 28-day mortality, delayed discontinuation of MV and prolonged ICU stay.

We aimed to evaluate cardiac function by trans-epicardial echocardiography in a pioneering ovine model of severe cardiopulmonary failure (CPF). Six healthy female sheep (weight 61±6.5 kg) were anesthetized, tracheally intubated and mechanically ventilated. An arterial catheter was placed for blood sampling and pressure monitoring. Left femoral artery and right jugular vein were cannulated and veno-arterial extracorporeal membrane oxygenation support commenced at a rate of 1.57±0.60 L/min. A left mini-thoracotomy was performed to access the pericardial sac. Heart failure was developed through intra-myocardial ethanol injections in the left ventricle. Ventilatory support was substantially decreased to concomitantly develop pulmonary failure. Cardiac function pre and post-development of CPF was evaluated by trans-epicardial echocardiography (Philips IE-33, North Ryde, Australia) and analysed by dedicated software (TOMTEC Imaging Systems GmbH, Germany). CPF caused a drop in mean arterial pressure and oxygenation, which resulted in an upsurge of lactate. Corresponding with a drop of mean arterial pressure, LV systolic function including left ventricular (LV) fractional area change, global circumferential strain and global radial strain also significantly decreased. In an innovative animal model of CPF, epicardial echocardiography is a reliable and feasible mean to follow-up dynamics of cardiac dysfunction.