Intravenous Autologous Bone Marrow-derived Mesenchymal Stromal Cells Delay Acute Respiratory Distress Syndrome in Swine.

Rationale: Early post injury mitigation strategies in ARDS are in short supply. Treatments with allogeneic stromal cells are administered after ARDS develops, require specialized expertise and equipment, and to date have shown limited benefit. Objectives: Assess the efficacy of immediate post injury intravenous administration of autologous or allogeneic bone marrow-derived mesenchymal stromal cells (MSCs) for the treatment of acute respiratory distress syndrome (ARDS) due to smoke inhalation and burns. Methods: Yorkshire swine (n = 32, 44.3 ± 0.5 kg) underwent intravenous anesthesia, placement of lines, severe smoke inhalation, and 40% total body surface area flame burns, followed by 72 hours of around-the-clock ICU care. Mechanical ventilation, fluids, pressors, bronchoscopic cast removal, daily lung computed tomography scans, and arterial blood assays were performed. After injury and 24 and 48 hours later, animals were randomized to receive autologous concentrated bone marrow aspirate (n = 10; 3 × 106 white blood cells and a mean of 56.6 × 106 platelets per dose), allogeneic MSCs (n = 10; 6.1 × 106 MSCs per dose) harvested from healthy donor swine, or no treatment in injured control animals (n = 12). Measurements and Main Results: The intravenous administration of MSCs after injury and at 24 and 48 hours delayed the onset of ARDS in swine treated with autologous MSCs (48 ± 10 h) versus control animals (14 ± 2 h) (P = 0.004), reduced ARDS severity at 24 (P < 0.001) and 48 (P = 0.003) hours, and demonstrated visibly diminished consolidation on computed tomography (not significant). Mortality at 72 hours was 1 in 10 (10%) in the autologous group, 5 in 10 (50%) in the allogeneic group, and 6 in 12 (50%) in injured control animals (not significant). Both autologous and allogeneic MSCs suppressed systemic concentrations of TNF-α (tumor necrosis factor-α). Conclusions: The intravenous administration of three doses of freshly processed autologous bone marrow-derived MSCs delays ARDS development and reduces its severity in swine. Bedside retrieval and administration of autologous MSCs in swine is feasible and may be a viable injury mitigation strategy for ARDS.

Effects of adjunct treatments on end-organ damage and histological injury severity in acute respiratory distress syndrome and multiorgan failure caused by smoke inhalation injury and burns.

BACKGROUND: We investigated effects of mesenchymal stem cells (MSC) or low-flow extracorporeal life support (ECLS) as adjunctive treatments for acute respiratory distress syndrome (ARDS) due to inhalation injury and burns. We hypothesized that these interventions decrease histological end-organ damage. METHODS: Anesthetized female swine underwent smoke inhalation injury and 40% TBSA burns, then critical care for 72h. The following groups were studied: CTR (no injury, n = 4), ICTR (injured untreated, n = 10), Allo (injured treated with allogenic MSC, n = 10), Auto (injured treated with autologous MSC, n = 10), Hemo (injured and treated with the Hemolung low flow ECLS system, n = 9), and Nova (injured and treated with the NovaLung low flow ECLS system, n = 8). Histology scores from lung, kidneys, liver, and jejunum were calculated. Data are presented as means±SEM. RESULTS: Survival at 72h was 100% in CTR; 40% in ICTR; 50% in Allo; 90% in Auto; 33% in Hemo; 63% in Nova. ARDS developed in 0/10 CTR; 10/10 ICTR; 8/9 Hemo; 5/8 Nova; 9/10 Allo; 6/10 Auto. Diffuse alveolar damage (DAD) was present in all injured groups. MSC groups had significantly lower DAD scores than ICTR animals (Allo 26.6 ± 3.4 and Auto 18.9 ± 1.5 vs. ICTR 46.8 ± 2.1, p < 0.001). MSC groups also had lower DAD scores than ECLS animals (Allo vs. Nova, p < 0.05, Allo vs. Hemo p < 0.001, Auto vs. Nova p < 0.001, Auto vs. Hemo, p < 0.001). Kidney injury ICTR (p < 0.05) and Hemo (p < 0.01) were higher than in CTR. By logistic regression, a PaO2-to-FiO2 ratio (PFR) < 300 was a function of the DAD score: logit (PFR < 300) = 0.84 + 0.072*DAD Score, odds ratio 1.074 (1.007, 1.147, p < 0.05) with a ROC AUC of 0.76, p < 0.001. CONCLUSION: Treatment with Auto MSC followed by Allo and then Nova were most effective in mitigating ARDS and MOF severity in this model. Further studies will elucidate the role of combination therapies of MSC and ECLS as comprehensive treatments for ARDS and MOF.

Dynamics of acute respiratory distress syndrome development due to smoke inhalation injury: Implications for prolonged field care.

BACKGROUND: Smoke inhalation injury (SII) causes 30% to 40% mortality and will increase as a cause of death during prolonged field care. We used a combat relevant model of acute respiratory distress syndrome due to SII to study temporal changes in ventilation-perfusion (V/Q) matching, computed tomography (CT) scan data, and histopathology and hypothesized that SII leads to increase in shunt (Qshunt), V/Q mismatch, lung consolidation, and diffuse alveolar damage. METHODS: Swine received severe SII and airway pressure release ventilation (APRV, n = 6), or conventional ARDSNet mechanical ventilation (MV) (CMV, n = 8). A control group without injury received volume controlled MV (CTRL, n = 6), The multiple inert gas elimination technique and CT were performed at baseline (BL), 0.5 hours, 1 hours, 2 hours, 24 hours, and 48 hours after injury. Diffuse alveolar damage scoring was performed post mortem. Significance at p less than 0.05: APRV versus CTRL; CMV versus CTRL; APRV versus CMV*; denotes changes versus BL. RESULTS: (1) SII caused increases in Qshunt more so in APRV than CMV group. Qshunt did not change in CTRL. (2) PaO2-to-FIO2 ratio (PFR) was lower in APRV versus CTRL at 2 hours (375 ± 62‡ vs. 549 ± 40) and 24 hours (126 ± 34‡* vs. 445 ± 5) and 48 hours (120 ± 41‡& vs. 430 ± 13). In CMV animals, PFR was lower versus CTRL and BL at 24 hours (238 ± 33) and 48 hours (98 ± 27). Qshunt correlated with PFR (r = 0.75, p < 0.0001, APRV and (r = 0.65, p < 0.0001, CMV). CT showed decrease in normally aerated lung, while poorly and nonaerated lung increased. CONCLUSION: Smoke inhalation injury leads to early development of shunt, V/Q mismatch, lung consolidation, and diffuse alveolar damage. These data substantiate the need for new point of injury interventions in the prolonged field care setting. LEVEL OF EVIDENCE: Animal research.

Point-of-care endoscopic optical coherence tomography detects changes in mucosal thickness in ARDS due to smoke inhalation and burns.

BACKGROUND: The prevalence of acute respiratory distress syndrome (ARDS) in mechanically ventilated burn patients is 33%, with mortality varying from 11-46% depending on ARDS severity. Despite the new Berlin definition for ARDS, prompt bedside diagnosis is lacking. We developed and tested a bedside technique of fiberoptic-bronchoscopy-based optical coherence tomography (OCT) measurement of airway mucosal thickness (MT) for diagnosis of ARDS following smoke inhalation injury (SII) and burns. METHODS: 16 female Yorkshire pigs received SII and 40% thermal burns. OCT MT and PaO2-to-FiO2 ratio (PFR) measurements were taken at baseline, after injury, and at 24, 48, and 72h after injury. RESULTS: Injury led to thickening of MT which was sustained in animals that developed ARDS. Significant correlations were found between MT, PFR, peak inspiratory pressure (PIP), and total infused fluid volume. CONCLUSIONS: OCT is a useful tool to quantify MT changes in the airway following SII and burns. OCT may be effective as a diagnostic tool in the early stages of SII-induced ARDS and should be tested in humans.

Resuscitative endovascular balloon occlusion of the aorta (REBOA): Comparison with immediate transfusion following massive hemorrhage in swine.

BACKGROUND: Resuscitative endovascular balloon occlusion of the aorta (REBOA) is less invasive than emergency department thoracotomy for the treatment of massive hemorrhage. We evaluated the effects of REBOA on carotid blood flow (Qcarotid) in a porcine model of massive hemorrhage. We hypothesized that REBOA restores Qcarotid faster than reinfusion of blood. METHODS: Spontaneously breathing sedated Sinclair pigs underwent exponential hemorrhage of 65% total blood volume in 1 hour. They were randomized into three groups. Positive control (PC, n = 7) underwent immediate transfusion of shed blood. REBOA (n = 21) received a novel 7 Fr ER-REBOA catheter (Pryor Medical, Arvada, CO) placed into aortic Zone 1 via a femoral artery introducer for 30 minutes or 60 minutes, with transfusion either after deflation or midway through inflation. Negative control (n = 7) received no resuscitation. Qcarotid was recorded continuously using an ultrasonic flow probe. Survival and time between Qcarotid, min and both a stable maximal value (Qcarotid, max) and restoration of baseline flow (Qcarotid, new BL) were compared by Kaplan-Meier analysis. RESULTS: Median time to Qcarotid, max was 3.0 minutes in the REBOA group versus 9.6 minutes in the control group (p = 0.006). Median time to Qcarotid, new BL was 6.0 minutes in the REBOA group versus 20.5 minutes in the PC group (p = 0.11). Slope of the linear regression between Qcarotid, min and Qcarotid, new BL was 16.7 in REBOA and 10.4 in PC (p = 0.31). Four-hour survival was 95% (20 of 21) in the REBOA group versus 71% (5 of 7) in the PC group (p = 0.06) and 0% in the negative control group. CONCLUSION: REBOA resulted in the restoration of Qcarotid ("cerebrovascular resuscitation") at least as rapidly as retransfusion of shed blood, with equivalent 4-hour survival. Further studies of REBOA, to include mitigation of end-organ effects and longer follow-up, are needed.

Comparison of virtual bronchoscopy to fiber-optic bronchoscopy for assessment of inhalation injury severity.

PURPOSE: Compare virtual bronchoscopy (VB) to fiberoptic bronchoscopy (FOB) for scoring smoke inhalation injury (SII). METHODS: Swine underwent computerized tomography (CT) with VB and FOB before (0) and 24 and 48 h after SII. VB and FOB images were scored by 5 providers off line. RESULTS: FOB and VB scores increased over time (p<0.001) with FOB scoring higher than VB at 0 (0.30±0.79 vs. 0.03±0.17), 24 h (4.21±1.68 vs. 2.47±1.50), and 48h (4.55±1.83 vs. 1.94±1.29). FOB and VB showed association with PaO2-to-FiO2 ratios (PFR) with areas under receiver operating characteristic curves (ROC): for PFR≤300, VB 0.830, FOB 0.863; for PFR≤200, VB 0.794, FOB 0.825; for PFR≤100, VB 0.747, FOB 0.777 (all p<0.001). FOB showed 80.3% specificity, 77% sensitivity, 88.8% negative-predictive value (NPV), and 62.3% positive-predictive value (PPV) for PFR≤300 and VB showed 67.2% specificity, 85.5% sensitivity, 91.3% NPV, and 53.4% PPV. CONCLUSIONS: VB provided similar injury severity scores to FOB, correlated with PFR, and reliably detected airway narrowing. VB performed during admission CT may be a useful screening tool specifically to demonstrate airway narrowing induced by SII.

Dynamic changes in shunt and ventilation-perfusion mismatch following experimental pulmonary contusion.

The objective of this study was to investigate early changes in oxygenation by means of the multiple inert gas elimination technique and in coagulation by means of thromboelastography (TEG) after right-sided pulmonary contusion (PC) in swine. Anesthetized swine (group 1; n = 8) sustained a right-chest PC by a captive-bolt stunner. Multiple inert gas elimination technique, TEG, and thoracic computed tomography (CT) scans were performed before and 10, 30, 60, and 120 min after injury. Three-dimensional CT scan reconstruction enabled measurement of volumes of poorly (Vol(Poor)) and nonaerated (Vol(Non)) lung. Eight animals (group 0) were used as uninjured controls. Pulmonary contusion led to sustained tachycardia and transient hypotension. Partial pressure of arterial oxygen (PaO2) decreased from 83.9 +/- 4.2 mmHg at baseline to 51.3 +/- 2.8 mmHg 10 min after PC (P < 0.001). Vol(Poor) and Vol(Non) on the right increased significantly after PC, followed by gradual progression in injury marked by decreased Vol(Poor) and increased Vol(Non). By the multiple inert gas elimination technique, blood flow to the true shunt compartment increased from 4.4% +/- 1.0% at baseline to 21.2% +/- 4.9% 10 min after PC, P < 0.001, peaked at 33.2% +/- 7.5% 30 min after PC, P < 0.001, and remained significantly higher compared with controls. Transient increase in blood flow to low and very low ventilation-perfusion (V/Q) compartments was also seen. Clot reaction time and formation rate by TEG increased at 2 h after PC. True shunt is the major cause of hypoxemia after PC, but V/Q mismatch also contributes significantly early after injury. By CT, PC leads to significant loss of functional lung volume on the side of injury. A mild hypocoagulable state was identified 2 h after injury.