Jay Steinberg

Altered alveolar mechanics in the acutely injured lung.

ObjectivesAlterations in alveolar mechanics (i.e., the dynamic change in alveolar size during tidal ventilation) are thought to play a critical role in acute lung injuries such as acute respiratory distress syndrome (ARDS). In this study, we describe and quantify the dynamic changes in alveolar mechanics of individual alveoli in a porcine ARDS model by direct visualization using in vivo microscopy. DesignProspective, observational, controlled study. SettingUniversity research laboratory. SubjectsTen adult pigs. InterventionsPigs were anesthetized and placed on mechanical ventilation, underwent a left thoracotomy, and were separated into the following two groups post hoc: a control group of instrumented animals with no lung injury (n = 5), and a lung injury group in which lung injury was induced by tracheal Tween instillation, causing surfactant deactivation (n = 5). Pulmonary and systemic hemodynamics, blood gases, lung pressures, subpleural blood flow (laser Doppler), and alveolar mechanics (in vivo microscopy) were measured in both groups. Alveolar size was measured at peak inspiration (I) and end expiration (E) on individual subpleural alveoli by image analysis. Histologic sections of lung tissue were taken at necropsy from the injury group. Measurements and Main Results In the acutely injured lung, three distinct alveolar inflation-deflation patterns were observed and classified: type I alveoli (n = 37) changed size minimally (I − E&Dgr; = 367 ± 88 &mgr;m2) during tidal ventilation; type II alveoli (n = 37) changed size dramatically (I − E&Dgr; = 9326 ± 1010 &mgr;m2) with tidal ventilation but did not totally collapse at end expiration; and type III alveoli (n = 12) demonstrated an even greater size change than did type II alveoli (I − E&Dgr; = 15,418 ± 1995 &mgr;m2), and were distinguished from type II in that they totally collapsed at end expiration (atelectasis) and reinflated during inspiration. We have termed the abnormal alveolar inflation pattern of type II and III alveoli “repetitive alveolar collapse and expansion” (RACE). RACE describes all alveoli that visibly change volume with ventilation, regardless of whether these alveoli collapse totally (type III) at end expiration. Thus, the term “collapse” in RACE refers to a visibly obvious collapse of the alveolus during expiration, whether this collapse is total or partial. In the normal lung, all alveoli measured exhibited type I mechanics. Alveoli were significantly larger at peak inspiration in type II (18,266 ± 1317 &mgr;m2, n = 37) and III (15,418 ± 1995 &mgr;m2, n = 12) alveoli as compared with type I (8214 ± 655 &mgr;m2, n = 37). Tween caused a heterogenous lung injury with areas of normal alveolar mechanics adjacent to areas of abnormal alveolar mechanics. Subsequent histologic sections from normal areas exhibited no pathology, whereas lung tissue from areas with RACE mechanics demonstrated alveolar collapse, atelectasis, and leukocyte infiltration. ConclusionAlveolar mechanics are altered in the acutely injured lung as demonstrated by the development of alveolar instability (RACE) and the increase in alveolar size at peak inspiration. Alveolar instability varied from alveolus to alveolus in the same microscopic field and included alveoli that changed area greatly with tidal ventilation but remained patent at end expiration and those that totally collapsed and reexpanded with each breath. Thus, alterations in alveolar mechanics in the acutely injured lung are complex, and attempts to assess what may be occurring at the alveolar level from analysis of inflection points on the whole-lung pressure/volume curve are likely to be erroneous. We speculate that the mechanism of ventilator-induced lung injury may involve altered alveolar mechanics, specifically RACE and alveolar overdistension.

Effect of positive end-expiratory pressure and tidal volume on lung injury induced by alveolar instability

IntroductionOne potential mechanism of ventilator-induced lung injury (VILI) is due to shear stresses associated with alveolar instability (recruitment/derecruitment). It has been postulated that the optimal combination of tidal volume (Vt) and positive end-expiratory pressure (PEEP) stabilizes alveoli, thus diminishing recruitment/derecruitment and reducing VILI. In this study we directly visualized the effect of Vt and PEEP on alveolar mechanics and correlated alveolar stability with lung injury.MethodsIn vivo microscopy was utilized in a surfactant deactivation porcine ARDS model to observe the effects of Vt and PEEP on alveolar mechanics. In phase I (n = 3), nine combinations of Vt and PEEP were evaluated to determine which combination resulted in the most and least alveolar instability. In phase II (n = 6), data from phase I were utilized to separate animals into two groups based on the combination of Vt and PEEP that caused the most alveolar stability (high Vt [15 cc/kg] plus low PEEP [5 cmH2O]) and least alveolar stability (low Vt [6 cc/kg] and plus PEEP [20 cmH2O]). The animals were ventilated for three hours following lung injury, with in vivo alveolar stability measured and VILI assessed by lung function, blood gases, morphometrically, and by changes in inflammatory mediators.ResultsHigh Vt/low PEEP resulted in the most alveolar instability and lung injury, as indicated by lung function and morphometric analysis of lung tissue. Low Vt/high PEEP stabilized alveoli, improved oxygenation, and reduced lung injury. There were no significant differences between groups in plasma or bronchoalveolar lavage cytokines or proteases.ConclusionA ventilatory strategy employing high Vt and low PEEP causes alveolar instability, and to our knowledge this is the first study to confirm this finding by direct visualization. These studies demonstrate that low Vt and high PEEP work synergistically to stabilize alveoli, although increased PEEP is more effective at stabilizing alveoli than reduced Vt. In this animal model of ARDS, alveolar instability results in lung injury (VILI) with minimal changes in plasma and bronchoalveolar lavage cytokines and proteases. This suggests that the mechanism of lung injury in the high Vt/low PEEP group was mechanical, not inflammatory in nature.

Alveolar inflation during generation of a quasi-static pressure/volume curve in the acutely injured lung.

ObjectiveLower and upper inflection points on the quasi-static curve representing a composite of pressure/volume from the whole lung are hypothesized to represent initial alveolar recruitment and overdistension, respectively, and are currently utilized to adjust mechanical ventilation in patients with acute respiratory distress syndrome. However, alveoli have never been directly observed during the generation of a pressure/volume curve to confirm this hypothesis. In this study, we visualized the inflation of individual alveoli during the generation of a pressure/volume curve by direct visualization using in vivo microscopy in a surfactant deactivation model of lung injury in pigs. DesignProspective, observational, controlled study. SettingUniversity research laboratory. SubjectsEight adult pigs. InterventionsPigs were anesthetized and administered mechanical ventilation, underwent a left thoracotomy, and were separated into two groups: control pigs (n = 3) were subjected to surgical intervention, and Tween lavage pigs (n = 5) were subjected to surgical intervention plus surfactant deactivation by Tween lavage (1.5 mL/kg 5% solution of Tween in saline). The microscope was then attached to the lung, and the size of each was alveolus quantified by measuring the alveolar area by computer image analysis. Each alveolus in the microscopic field was assigned to one of three types, based on alveolar mechanics: type I, no visible change in alveolar size during ventilation; type II, alveoli visibly change size during ventilation but do not totally collapse at end expiration; and type III, alveoli visibly change size during tidal ventilation and completely collapse at end expiration. After alveolar classification, the animals were disconnected from the ventilator and attached to a super syringe filled with 100% oxygen. The lung was inflated from 0 to 220 mL in 20-mL increments with a 10-sec pause between increments for airway pressure and alveolar confirmation to stabilize. These data were utilized to generate both quasi-static pressure/volume curves and individual alveolar pressure/area curves. Measurements and Main ResultsThe normal lung quasi-static pressure/volume curve has a single lower inflection point, whereas the curve after Tween has an inflection point at 8 mm Hg and a second at 24 mm Hg. Normal alveoli in the control group are all type I and do not change size appreciably during generation of the quasi-static pressure/volume curve. Surfactant deactivation causes a heterogenous injury, with all three alveolar types present in the same microscopic field. The inflation pattern of each alveolar type after surfactant deactivation by Tween was notably different. Type I alveoli in either the control or Tween group demonstrated minimal change in alveolar area with lung inflation. Type I alveolar area was significantly (p < .05) larger in the control as compared with the Tween group. In the Tween group, type II alveoli increased significantly in area, with lung inflation from 0 mL (9666 ± 1340 &mgr;m2) to 40 mL (12,935 ± 1725 &mgr;m2) but did not increase further (220 mL, 14,058 ± 1740 &mgr;m2) with lung inflation. Type III alveoli initially recruited with a relatively small area (20 mL lung volume, 798 ± 797 &mgr;m2) and progressively increased in area throughout lung inflation (120 mL, 7302 ± 1405 &mgr;m2; 220 mL, 11,460 ± 1078 &mgr;m2) ConclusionThe normal lung does not increase in volume by simple isotropic (balloon-like) expansion of alveoli, as evidenced by the horizontal (no change in alveolar area with increases in airway pressure) pressure/area curve. After surfactant deactivation, the alveolar inflation pattern becomes very complex, with each alveolar type (I, II, and III) displaying a distinct pattern. None of the alveolar pressure/area curves directly parallel the quasi-static lung pressure/volume curve. Of the 16, only one type III atelectatic alveolus recruited at the first inflection point and only five recruited concomitant with the second inflation point, suggesting that neither inflection point was due to massive alveolar recruitment. Thus, the components responsible for the shape of the pressure/volume curve include all of the individual alveolar pressure/area curves, plus changes in alveolar duct and airway size, and the elastic forces in the pulmonary parenchyma and the chest wall.

Metalloproteinase inhibition reduces lung injury and improves survival after cecal ligation and puncture in rats

BACKGROUND Neutrophil activation with concomitant matrix metalloproteinase-2 (MMP-2) and matrix metalloproteinase-9 (MMP-9) release has been implicated in the development of sepsis-induced acute lung injury. We hypothesized that COL-3, a chemically modified tetracycline known to inhibit MMP-2 and MMP-9, would reduce lung injury and improve survival in rats following cecal ligation and puncture (CLP). METHODS Sprague-Dawley rats were separated into five groups: 1) sham CLP+ carboxymethylcellulose (CMC; vehicle for COL-3, n = 6); 2) sham CLP + COL-3 (n = 6); 3) CLP + CMC (n = 10); 4) CLP + single-dose (SD) COL-3 administered concomitant with CLP (n = 9); and 5) CLP + multiple-dose (MD) COL-3 administered concomitant with CLP and at 24 h after CLP (n = 15). Rats were sacrificed at 168 h (7 days) or immediately after death, with survival defined as hours after CLP. Histological lung assessment was made based on neutrophil infiltration, alveolar wall thickening, and intraalveolar edema fluid. Lung MMP-2 and MMP-9 levels were assessed by immunohistochemistry. MMP-2 and MMP-9 levels were correlated with survival by simple regression analysis. RESULTS The mortality of rats in the cecal ligation and puncture without treatment group (CLP + CMC) was 70% at 168 h. A single dose of COL-3 in the CLP + COL-3 (SD) group significantly reduced mortality to 54%. Furthermore, with a repeat dose of COL-3 at 24 h after CLP, mortality was significantly reduced to 33%. Pathologic lung changes seen histologically in the CLP + CMC group were significantly reduced by COL-3. A significant reduction in lung tissue levels of MMP-2 and MMP-9 was noted in both groups treated with COL-3. Reduction of MMP-2 and MMP-9 levels correlated with improved survival. CONCLUSION Inhibition of MMP-2 and MMP-9 by COL-3 in a clinically relevant model of sepsis-induced acute lung injury reduces pulmonary injury and improves survival in a dose-dependent fashion. Our results suggest that prophylactic treatment with COL-3 in high-risk patients may reduce the morbidity and mortality associated with sepsis-induced acute respiratory distress syndrome.

Chemically modified tetracycline prevents the development of septic shock and acute respiratory distress syndrome in a clinically applicable porcine model.

Sepsis causes more than with 215,000 deaths per year in the United States alone. Death can be caused by multiple system organ failure, with the lung, in the form of the acute respiratory distress syndrome (ARDS), often being the first organ to fail. We developed a chronic porcine model of septic shock and ARDS and hypothesized that blocking the proteases neutrophil elastase (NE) and matrix metalloproteinases (MMP-2 and MMP-9) with the modified tetracycline, COL-3, would significantly improve morbidity in this model. Pigs were anesthetized and instrumented for hemodynamic monitoring and were then randomized to one of three groups: control (n = 3), laparotomy only; superior mesenteric artery occlusion (SMA) + fecal blood clot (FC; n = 7), with intraperitoneal placement of a FC; and SMA + FC + COL (n = 5), ingestion of COL-3 12 h before injury. Animals emerged from anesthesia and were monitored and treated with fluids and antibiotics in an animal intensive care unit continuously for 48 h. Serum and bronchoalveolar lavage fluid (BALF) were sampled and bacterial cultures, MMP-2, MMP-9, NE, and multiple cytokine concentrations were measured. Pigs were reanesthetized and placed on a ventilator when significant lung impairment occurred (PaO2/FiO2 < 250). At necropsy, lung water and histology were assessed. All animals in the SMA + FC group developed septic shock evidenced by a significant fall in arterial blood pressure that was not responsive to fluids. Lung injury typical of ARDS (i.e., a fall in lung compliance and PaO2/FiO2 ratio and a significant increase in lung water) developed in this group. Additionally, there was a significant increase in plasma IL-1 and IL-6 and in BALF IL-6, IL-8, IL-10, NE, and protein concentration in the SMA + FC group. COL-3 treatment prevented septic shock and ARDS and significantly decreased cytokine levels in plasma and BALF. COL-3 treatment also significantly reduced NE activity (P < 0.05) and reduced MMP-2 and MMP-9 activity in BALF by 64% and 34%, respectively, compared with the SMA + FC group. We conclude that prophylactic COL-3 prevented the development of ARDS and unexpectedly also prevented septic shock in a chronic insidious onset animal model of sepsis-induced ARDS. The mechanism of this protection is unclear, as COL-3 inhibited numerous inflammatory mediators. Nevertheless, COL-3 significantly reduced the morbidity in a clinically applicable animal model, demonstrating the possibility that COL-3 may be useful in reducing the morbidity associated with sepsis and ischemia/reperfusion injury in patients.

The Development Of Acute Respiratory Distress Syndrome After Gut Ischemia/reperfusion Injury Followed By Fecal Peritonitis In Pigs: A Clinically Relevant Model

Numerous clinical trials using anti-inflammatory agents for patients with acute respiratory distress syndrome (ARDS) have failed despite efficacy in acute animal models. This underscores the necessity of developing a clinically relevant model of ARDS. Initially, we attempted to induce lung injury in pigs by fecal peritonitis only. When this was unsuccessful, we designed a two-hit model of ischemia/reperfusion (I/R) injury followed by fecal peritonitis to create a clinically applicable model of ARDS. The initial study consisted of Yorkshire swine [group 1, fecal clot (FC), n = 4] that were followed clinically after intraperitoneal placement of a fecal (0.5 mL/kg) blood (2 mL/kg) clot. Blood was sampled daily for cultures, a complete blood count, a lactate level, and various cytokine expression determined by enzyme-linked immunosorbent assay (ELISA). Pigs were treated with antibiotics and fluids, placed on a ventilator before sacrifice to obtain hemodynamic and pulmonary parameters, and underwent histologic lung assessment. Additionally, bronchoalveolar lavage fluid was obtained for protein concentration and cytokine levels. Once it was evident that no lung injury had occurred, we designed a more severe model. A second group of Yorkshire swine [group 2, superior mesenteric artery (SMA) + FC, n = 4] underwent SMA occlusion for 30 min (I/R) followed by intraperitoneal placement of a FC as in the initial group. These pigs were monitored more invasively and continuously in an intensive care setting for 48 h and followed, treated, and assessed in a similar fashion to group 1. Group 1 (FC) pigs survived 9 days and showed signs of sepsis (bacteremia with polymicrobial organisms), an inflammatory response in the form of elevated cytokines, yet no physiologic or histologic evidence of lung injury. Group 2 (SMA + FC) pigs demonstrated more severe sepsis, a significantly increased cytokine response compared with animals in the FC group, and physiologic signs of progressive pulmonary injury. Pigs in the SMA + FC group were sacrificed at 48 h after clinical deterioration (significant decline in oxygenation) and demonstrated pathologic evidence of lung injury indicated by increased bronchoalveolar lavage fluid protein, diffuse and thickened alveolar septae, hyaline membrane formation, and pulmonary edema. The addition of a second “hit” (SMA occlusion, I/R) to a FC sepsis model resulted in severe lung injury that developed within a 3-day period. To our knowledge, this is the first large animal experiment that definitively and consistently causes insidious onset ARDS in pigs. By closely paralleling the clinical development of pulmonary injury, this model should prove invaluable in the study of human ARDS.

New approach to the surgical management of pulmonary arteriovenous malformations after cavopulmonary anastomosis

The development of pulmonary arteriovenous malformations after cavopulmonary bypass in patients with congenital heart disease is well documented. We report successful management of pulmonary arteriovenous malformations after cavopulmonary bypass in a patient with an interrupted inferior vena cava (IVC) and multiple hepatic veins utilizing an extracardiac conduit from the hepatic veins to the hemiazygous continuation of the interrupted IVC. This technique, performed without circulatory arrest or an atriotomy, may limit morbidity associated with intracardiac procedures in patients with single ventricle morphology. Furthermore, this case suggests an alternative technique for completion Fontan in patients with an interrupted IVC and multiple hepatic venous drainage.

Alveolar mechanics alter hypoxic pulmonary vasoconstriction.

ObjectivesHypoxic pulmonary vasoconstriction is the primary physiologic mechanism that maintains a proper ventilation/perfusion match, but it fails in diffuse lung injuries such as acute respiratory distress syndrome. Acute respiratory distress syndrome is associated with pulmonary surfactant loss that alters alveolar mechanics (i.e., dynamic change in alveolar size and shape during ventilation), converting normal stable alveoli into unstable alveoli. We hypothesized that alveolar instability stents open pulmonary microvessels and is the mechanism of hypoxic pulmonary vasoconstriction failure associated with acute respiratory distress syndrome. DesignProspective, randomized, controlled study. SettingUniversity research laboratory. SubjectsTen adult pigs. InterventionsAnesthetized ventilated pigs were prepared surgically for hemodynamic monitoring and were subjected to a right thoracotomy. An in vivo microscope was attached to the right lung, and the microvascular response to hypoxia (Fio2, 15%) was measured in a lung with normal stable alveoli and in a lung with unstable alveoli caused by surfactant deactivation (Tween lavage). Measurements and Main ResultsAlveolar instability, defined as the difference between alveolar area at peak inspiration and end expiration and assessed as a percentage change (I-E&Dgr;%), was significantly increased after Tween (23.9 ± 3.0, I-E&Dgr;%) compared with baseline (2.4 ± 1.0, I-E&Dgr;%). Alveolar instability was associated with the following microvascular changes: a) increased vasoconstriction (Tween, 14.9 ± 1.0%) in response to hypoxia compared with baseline (10.8 ± 1.2%, p < .05); and b) increased mean vascular diameter (Tween, 41.2 ± 1.5 &mgr;m) compared with the mean diameter at baseline (24.6 ± 1.0 &mgr;m, p < .05). ConclusionUnstable alveoli stent open pulmonary vessels, which may explain the failure of hypoxic pulmonary vasoconstriction in acute respiratory distress syndrome.

Chemically Modified Tetracycline Improves Contractility in Porcine Coronary Ischemia/Reperfusion Injury

Abstract  Background: Reperfusion of ischemic myocardium has been implicated in extension of infarct size and deleterious clinical outcomes. Anti‐inflammatory agents reduce this reperfusion injury. Chemically modified tetracycline‐3 (CMT‐3) (Collagenex Pharmaceuticals, Newtown, PA, USA) lacks antimicrobial properties yet retains anti‐inflammatory activity. We examined infarct size and myocardial function in a porcine coronary artery occlusion/reperfusion model in CMT‐3‐treated and control animals. Methods: Yorkshire pigs (n = 8) underwent median sternotomy, pretreatment with heparin (300 U/kg and 67 U/kg/hr IV) and lidocaine (1 mg/kg IV) and were divided into two groups. Group one (n = 4) had the left anterior descending artery (LAD) occluded for 1 hour, after which it was reperfused for 2 hours. Group two (n = 4) had an identical protocol to group one except CMT‐3 (2 mg/kg IV) was administered prior to occlusion of the LAD. Results: Animals receiving CMT‐3 had significantly decreased infarct size in relation to the ventricular area‐at‐risk (AAR) (28 ± 9% vs. 64 ± 8%; p < 0.05). Myocardial contractile function was superior in the CMT‐3 treatment, indicated by a higher cardiac index (2.9 ± 0.3 vs. 2.0 ± 0.3 L/min/m2; p < 0.05) and stroke volume index (22 ± 2 vs. 17 ± 1 L/m2/beat; p < 0.05). Conclusions: CMT‐3 decreased infarct size in relation to the AAR resulting in relative preservation of contractility, suggesting CMT‐3 may improve outcomes during myocardial ischemia reperfusion.