Efficacy and Safety of a Lung Recruitment Protocol in Children With Acute Lung Injury

I. Introduction A. Background Lung units which participate in gas exchange are known as 'recruited' lung. Patients with lung injury suffer from a proportion of units which do not participate in gas exchange (i.e. the derecruited state), at times resulting in impaired gas exchange. Derecruitment of alveoli may also cause intrapulmonary shunting and worsen lung injury through atelectotrauma7. Outcomes in acute respiratory distress syndrome have improved significantly Is this really true? since clinicians have begun to employ lung protective strategies, including low-tidal volume ventilation and permissive hypercapnea8, 9. However, low-tidal volume ventilation has been recognized to decrease recruited lung volume, a phenomenon which persists despite the aggressive positive end-expiratory pressure (PEEP) strategy employed in ARDSNet studies4. Atelectasis associated with low-tidal volume ventilation is relieved through the use of so-called sign breaths, or recruitment breaths10. Further, the proportion of lung remaining in the derecruited state may contribute to the morbidity and mortality associated with acute respiratory distress syndrome (ARDS)11. In adults, several strategies have been utilized to recruit the lung: sustained inflation (SI) and the maximal recruitment strategy. The so-called open lung approach (OLA) includes an SI followed by the setting of PEEP to the measured lower inflection point of the PV curve. An alternative approach to setting PEEP is a decremental PEEP titration, which includes the sequential lowering of PEEP until a predetermined decrement in PaO2 or SaO2 occurs. Studies which have not included a strategy for maintaining lung recruitment following a recruitment maneuver have all been studied. The impact of lung recruitment in the long-term course of ARDS is not yet clear. It is clear that lung recruitment is most effective earlier in the course of ALI/ARDS. Grasso et al demonstrated that patients who received a recruitment maneuver on day 1±0.3 of ARDS could be recruited, versus patients recruited on day 7±1. Similarly, Gattinoni et al11 and Crotti et al5 found limited recruitment in patients who were well along in the course of ARDS. Borges et al,6 Tugrul et al,12 and Girgis et al all recruited patients early in the course of ARDS, and each found marked lung recruitment, on average, in all the patients studied. Each of these studied demonstrated an ability to improve oxygen saturations and (sometimes studied) end-expiratory lung volume. While no study has examined the effect of this change on morbidity or mortality, in children hypoxemia is known to be a common cause of morbidity. Importantly in children, treatment of hypoxia often drives escalating ventilator settings, the use of high frequency oscillatory ventilation (HFOV) or the use of extra-corporeal membrane oxygenation (ECMO). Early recruitment in children with ALI/ARDS may prevent the need for escalation of care towards these more invasive, and risk-imposing therapies. Currently, there are no clinical practice guidelines or standard of care regarding lung recruitment strategies for children with lung injury. To date, no studies have been performed in children documenting the efficacy or safety of any of the strategies utilized in intensive care. The primary aim of this study is to demonstrate the efficacy and safety of sustained inflation and maximal recruitment maneuvers in children with lung injury. B. Lung Recruitment Maneuvers 1. Sustained Inflations (SI) Sustained inflations (SI) are commonly utilized in the intensive care unit as a recruitment maneuver. In SI, the patient is given a prolonged positive pressure breath (usually between 30 and 45 cm H2O) lasting between 15 and 40 seconds. Clinically, this maneuver is applied following a derecruitment, such as suctioning, or when the patient exhibits hypoxemia. This strategy is currently employed in the Children's Hospital Boston Medical-Surgical Intensive Care Unit in some patients with lung injury, though there is no published evidence of its safety or efficacy in children. Could we put the 3 SI studies in premature infants that use 20-40 cmH20 followed by CPAP? 1 was a rabbit study. The studies described below regard sustained inflation recruitment maneuvers. 1. Tugrul, et al12 studied the effect of SI in 24 adults with ARDS. SI included 45 cm H2O for 30 seconds, followed by a decremental PEEP titration from 20 cm H2O down, titrating to saturations. P/F ratio increased \>200 in 11/24 patients, which persisted at 6 hours post-maneuver in 9/24 patients. Barotrauma was not observed in any patients. 2. Richard, et al studied4 the effect of SI (45 cm H2O for 15 seconds) in 10 adults with ARDS ventilated with a low-tidal volume strategy. The maneuver induced a significant increase in recruited lung volumes (175±108 ml pre-maneuver versus 254±137 ml post-maneuver). This improvement was associated with an increase in arterial oxygen saturations. There was no report of sustainability of recruitment, nor of adverse effects. 3. Lapinski, et al13 studied the safety and efficacy of the SI maneuver in 14 adults with ARDS. The pressure utilized in the maneuver was the lesser of 45 cm H2O or the plateau pressure utilized in 12 mL/kg tidal volume breaths; the maneuver was held for 20 seconds. Oxygen saturations increased from 86.9 ± 5.5 to 94.3 ± 2.3% by 10 minutes, which was maintained at 4 hours in 10/14 patients. The four in whom saturations fell had PEEP levels below 10 cm H2O. The systolic blood pressure decreased during the 20 second inflation in some patients, mean change of 6.9 mm Hg, which reversed rapidly following release of the maneuver in all patients. No barotrauma was noted in any patient at 24 hour followup. 4. Pelosi, et al10 studied the effect of three consecutive 'sigh' breaths (one with plateau pressure of 45 cm H2O but of normal duration) in 10 adults with ARDS. After 1 hour of 3 sigh breaths per minute, end-expiratory lung volume (measured by He dilution) increased from 1.49±0.58 to 1.91±0.67 L, and an increase in PaO2 from 92.8±18.6 to 137.6±23.9 mm Hg, compared with the same ventilator settings without sigh breaths. Lung elastance and ventilation also increased. No adverse effects were noted. 5. Toth, et al studied the cardiac and respiratory changes which occur in SI (40 cm H2O for 40 seconds) in 18 adults with ARDS. The PaO2 increased from pre-recruiting maneuver to following the 60 minute followup period (203±108 vs. 322±101 mm Hg, p \< .001). Cardiac index (CI) and the intrathoracic blood volume (ITBV) decreased following the recruiting maneuver (CI, 3.90±1.04 vs. 3.62±0.91 L/min/m2, p \< .05; ITBVI, 832±205 vs. 795±188 mL/m2, p \< .05). There was no correlation with CI and mean arterial pressure, and no significant changes occurred in mean arterial blood pressure. 6. In perhaps the largest study of sustained inflations in the protective ventilation era, Meade, et al14 studied 983 adults with ARDS utilizing lung protective strategies (tidal volumes 6 mL/kg) in both groups. The experimental group utilized SI recruitment maneuvers (40 cm H2O for 40 seconds), higher levels of PEEP (14.6±3.4 in the so-called 'lung open ventilation' group versus 9.8±2.7 in the control group), and higher plateau pressures (30.2±6.3 versus 24.9±5.1 cm H2O) than the control group, which did not utilize recruitment maneuvers. While there was no difference in all-cause mortality, the experimental group had lower rates of refractory hypoxemia (4.6% vs 10.2%; RR, 0.54), death with refractory hypoxemia (4.2% vs 8.9%; RR, 0.56), and less frequent use of escalated therapies (defined as high frequency ventilation, inhaled nitric oxide, jet ventilation, or extra-corporeal support) (5.1% vs 9.3%; RR, 0.6). This lends credence to the notion that effective recruitment of children early in acute lung injury may avert escalation of therapy and decrease the mortality due to severe hypoxia. 2. Pressure control ventilation (PCV) recruitment maneuver In contrast to a SI, ventilating patients with increased plateau pressures and increased PEEP levels has also been used as a recruiting maneuver. In this case, the patient is temporarily placed on higher ventilator pressures than would be used to ventilate a patient otherwise. A combination of PEEP and plateau pressures helps to recruit atelectatic alveoli. Pressure control ventilation is the mode of choice in such patients because in it the plateau pressure is set, and therefore kept constant (in contrast to volume controlled ventilation). Thus, a known plateau pressure (which participates in lung recruitment) is delivered. Following are the salient studies utilizing a PCV recruitment maneuver. ? Athanasios "TLC maneuver" in anesthetized children? Small study treating intraoperative atelectisis. 1. Borges, Amato, et al6 have performed the most convincing and complete study of the maximal recruitment maneuver to date. In it, they studied 26 adults early in the course of ARDS comparing the open lung approach (including SI of 40 cm H2O for 40 seconds, followed by PEEP set at Pflex + 2 cm H2O, described below) with the maximal recruitment strategy, performed in sequence. Maximal recruitment was defined as a PaO2 plus PaCO2 exceeding 400 mm Hg, indicative of minimal intrapulmonary shunting. The strategy included pressure controlled ventilation of 15 cm H2O, with PEEP increased in 5 cm H2O increments every 5 minutes (maximum PEEP 45 cm H2O) until maximal recruitment was achieved. Maximal recruitment was achieved in 24/26 patients, two thirds of which were achieved by a PEEP of 30 cm H2O. Utilizing a PEEP titration to maintain lung recruitment (see below), PaO2 remained above 400 (lungs fully recruited) in all patients at 6 hours. The procedure was tolerated hemodynamically in all patients and the long-term incidence of barotrauma in study patients (7.7%) was lower than the recognized incidence in ARDSNet studies (10-11%)8. 2. Crotti, et al5 studied five adults with ALI/ARDS with varying plauteau pressures (10, 15, 20, 30, 35, 40, 45 cm H2O) and PEEP levels (5, 10, 15, 20 cm H2O). End-inspiratory and end-expiratory thoracic CT imaging was obtained with each combination of plateau pressure and PEEP. Percentage of lung recruited (as determined radiographically) increased in a nearly linear fashion with increasing plateau pressures (Figure 1). The maximal opening frequency occurred at 20 cm H2O, suggesting that plateau pressures exceeding 20 cm H2O would be most beneficial in lung recruitment. No measures of gas exchange or adverse events were studied. 3. Foti, et al15 compared changes in oxygenation and end-expiratory lung volumes at increasing PEEP levels with superimposed volume controlled ventilation versus a sustained inflation maneuver in 15 adults with ARDS. Although recruitment was variable (due to the use of VCV, hence variable plateau pressures), he found that maximal increase in oxygenation occurred at high sustained levels of PEEP (mean 16 cm H2O for 30 minutes) when compared with 30 second recruitment maneuvers and low PEEP (mean of 9 cm H2O). The PEEP/plateau protocol was not associated with any hemodynamic alterations, as with the sustained inflation maneuver. 4. Villagra, et al, studied a recruiting maneuver utilizing 2 minutes of ventilation using pressure-controlled ventilation (PCV) in 18 adults with early ARDS. For the RM, PEEP was set 3 cm H2O above the upper inflection point of the PV curve, and peak pressures set to 50. Importantly, RM peak pressures did not differ significantly from baseline ventilation (43.6±7.6 pre-RM, 47±4.5 during RM, and 42.8±7.1 cm H2O post-RM), although PEEP did (14±1.3 pre-RM vs 30±4.9 cm H2O). The investigators did not find a significant difference in PaO2 following the recruiting maneuver (127±63 pre-RM, 162±108 during, and 149±85 cm H2O post-RM). This finding implies several important findings regarding lung recruitment. First, the peak pressure represents a significant determinant in lung recruitment. By contrast, Borges, et al were able to fully recruit all 18 ARDS patients by using driving pressures differed significantly from pre-RM in the maximal recruitment phase (50-60 cm H2O during recruitment vs 30 cm H2O at baseline). Further, the recruiting maneuver utilized here was only 2 minutes, versus the incremental titration utilized by Borges which lasted 4 minutes in each phase. While not systematically studied, the lack of significant improvement in PaO2 post-RM found in this study suggests that recruitment utilizing PCV may be a function of time and peak pressures utilized in the RM. 5. Medoff,3 et al presented a single patient with refractory hypoxemia. Management included repeated sustained inflations with PEEP set just above Pflex based on the PV curve. A recruiting maneuver including 20 cm H2O of PEEP and 40 cm H2O distending pressure for 2 minutes was performed. The patient exhibited marked improvement in oxygenation and near-complete recruitment of the lung by CT scan (see Figure 2). 6. While all of the aforementioned studies achieved successful recruitment, Gattinoni, et al11 studied 68 adults with late ARDS (5±6 days ventilation at enrollment) who failed to successfully recruit. Recruitment maneuvers included pressure-controlled ventilation with peak pressure of 45 cm H2O, PEEP of 5 cm H2O, and respiratory rate of 10. PEEP levels were subsequently applied in random order, between 5 and 15 cm H2O. He found that study patients had only 13±11% recruitable lung by CT scan. Amongst all patients, he found that 24% of lung tissue was non-recruitable, even at airway pressures of 45 cm H2O. No markers of gas exchange were measured. It is important to note three aspects of this study which are likely responsible for the excessively high percentage of non-recruitable lung, in contrast with the highly successful recruitment strategy outlined by Borges, et al. First, Borges studied patients early in the ARDS course (median, 2 days) while Gattinoni studied late ARDS. Second, Borges utilized peak pressures as high as 60 cm H2O, while Gattinoni utilized pressures up to 45 cm H2O. Finally, Gattinoni did not practice a PEEP titration, but lowered PEEP to 5 cm H2O following recruitment maneuvers. Thus, Gattinoni's study highlights the importance of high peak pressures in lung recruitment, of recruiting early in ARDS, and of a decremental PEEP titration to maintain lung recruitment. C. Strategies of Maintaining Recruitment The application of positive end-expiratory pressure is known to prevent repetitive derecruitment-recruitment stresses on lung tissue. In an animal model of acute lung injury, Farias showed histologic and biochemical evidence of atelectotrauma is averted when PEEP is applied following lung recruitment7. The two principal strategies to prevent derecruitment following a recruitment maneuver include the open lung approach and decremental PEEP titration. 1. Open Lung Approach In the open lung approach, PEEP is set just above the lower inflection point (also known as Pflex) of the patient's pressure-volume curve. This theoretically prevents the PEEP from decreasing below the zone of underdistension. 1. In a second part of the study mentioned above, Richard, et al4 also studied the effect of increasing PEEP relative to the measured LIP (LIP + 4 cm H2O, no SI) compared to PEEP set at the lower inflection point following SI. In 10 adults ventilated with low-tidal volume ventilation, measured lung volumes were 175±108 at PEEPLIP versus 332±91 ml at PEEPLIP+4. This increase was even more notable in patients ventilated with high tidal volumes (10 mL/kg), demonstrating the role of plateau pressures (in addition to PEEP) in alveolar recruitment (see Figure 3). 2. Amato, et al16 studied the effects of a 'protective ventilation strategy' in 53 patients with ARDS. The protective ventilation group received recruitment maneuvers via sustained inflation (35-40 cm H2O for 40 seconds) followed by PEEP set at LIP + 2 cm H2O. The control group received a PEEP titrated to saturations, no recruiting maneuvers. The protective ventilation group had a dramatic decrease in 28 day all-cause mortality (38% vs 71%, p\<0.001). However, the control group also received high tidal volume ventilation (12 mL/kg), later proven to be a significant determinant in ARDS mortality.8 Therefore, the contribution of RM and open lung approach to the improved mortality of this study, if any, is difficult to conclude. 2. Decremental PEEP Titration In a decremental PEEP titration, following a recruiting maneuver, the PEEP is set at a high level, often between 20 and 26 cm H2O. PEEP is decreased incrementally, and markers of lung inflation (e.g. gas exchange or measured lung volume) are followed. When evidence of atelectasis occurs, PEEP is held at or just above this level, often termed 'optimal PEEP.' 1. Borges, Amato, et al6 describe a PEEP titration following a maximal recruitment protocol and compare it to the open lung approach (PEEP set at Pflex + 2 cm H2O). Following a maximal recruitment strategy (described above), PEEP was set at 25 cm H2O and decreased by 2 cm H2O every 4 minutes until PaO2 + PaCO2 \< 380 mm Hg, then increased by 2 cm H2O (optimal PEEP, mean 20±5 cm H2O). Using this strategy, PaO2 was sustained at levels above 400 mm Hg at 30 minute followup in all 24 patients (see Figure 4, top), and above 450 in the 16 patients followed for 6 hours. This correlated with a percent mass of collapsed tissue as measured by CT scan of \<5% at end of PEEP titration and at 30 minute followup (see Figure 4, bottom). 2. Toth, et al17 described a decremental PEEP titration following a SI recruiting maneuver. Following SI at 40 cm H2O for 40 seconds, PEEP was set at 26 cm H2O, then decreased by 2 cm H2O every 4 minutes until the PaO2 decreased by \>10% of its peak value. The PEEP 2 cm H2O above this was defined as optimal PEEP. Optimal PEEP was found to actually be lower than the patients' baseline PEEP (prior to protocol initiation, 17±3 vs 15±4 cm H2O). The PaO2 following the recruiting maneuver (203±108 at baseline PEEP vs 328±132 cm H2O) sustained at the 30 minute followup period (266±121 cm H2O, p\<0.05 compared to baseline). D. Estimating Lung Volumes Currently, the ability to determine whether a patient is at, below or above their ideal functional residual capacity is deduced from surrogate measurements, including lung appearance of the chest radiograph, vital sign trends (particularly oxygenation), and pressure-volume (P-V) curves generated by modern ventilators. In this study, we will utilize an established method of determining lung volume (MBNW) to study lung volumes. 1. Pressure-Volume Curves The pressure-volume curve represents the continual relationship between changes in pressure and changes in volume of the lung. The slope of the line represents the compliance of the lung. In Figure 5, note the three lines which comprise the inspiratory limb (lower curve). The leftmost line represents noncompliant, atelectatic lung. The point at which the slope changes is known as the lower infection point, also known as Pflex. In the open lung approach (discussed later), PEEP is set to a pressure just above the lower inflection point. Physiologically, it is hypothesized that this is the point at which all atelectatic segments of lung are recruited, and that disallowing ventilator pressures from dropping below this at any point (by setting PEEP above this level) minimizes atelectasis. The point which distinguishes the second change in slope of the line (becoming flat again) is known as the upper inflection point (UIP). Pressures above this point represent overdistended, noncompliant alveoli, and thus this point represents the pressure at which the compliance of the lung decreases dramatically. In this protocol, the PV curve will be measured for each patient and the UIP utilized as the ceiling pressure at any point in the protocol. In this way, we will recruit compliant areas of the lung without the risks of overdistension. 2. Nitrogen Multiple Breath Washout Technique (MBNW) MBNW has been utilized in a number of clinical studies, and is considered to be a gold standard in the measurement of lung volume18-21. Currently, the most accurate way to measure the volume of the lung is through dilution of a known amount of a gas with low solubility be rebreathing in a closed system. The changes in concentration with sequential breaths allow a calculation of the volume of distribution of the gas. One gas which has been utilized for this purpose is nitrogen22, appealing due to its ubiquitous presence in the environment. Measurement of nitrogen gas concentrations, however, is available only using gas chromatography or mass spectrometry, neither of which is clinically practical. Recently, a technique has been validated by which the partial pressure of nitrogen is calculated as the residual of partial pressures of oxygen gas, carbon dioxide gas and nitrogen gas, which together comprise the only three important gases in a ventilator circuit. The former two gases are readily measured in a ventilator circuit in real time, but of course vary widely with the metabolic state of the patient. Stenqvist has developed the NMBW technique to calculate FRC using the changes in exhaled O2 and CO2, manipulating inspired oxygen concentration to alter fraction of inspired nitrogen23. FRC is calculated as follows: FiN2 = 1-FiO2 (set by ventilator) FeO2 = 1-FeO2 (measured)-FeCO2 (measured) Inspired and expired alveolar tidal volumes are calculated using O2 consumption (VO2) and CO2 production (VCO2) as calculated by indirect calorimetry24: Volumes of inspired and expired nitrogen gas associated with a single breath are calculated from end-tidal nitrogen content (EtN2, inferred from measured expired CO2 and oxygen content), inspired nitrogen fraction (FiN2) and inspiratory and expiratory alveolar tidal volume as follows: Before making the incremental 10% change in FiN2 via manipulating FiO2, baseline values for VO2, VCO2 and ETN2 are made. VO2 and VCO2 are assumed to be constant throughout the measurement. The FiN2 is then manipulated, and FRC estimated as follows: Measurement of FRC using this methodology in a lung model of known oxygen consumption and lung volumes23 revealed excellent precision (mean FRC 103 5%) even when utilizing incremental changes in FiO2 from 0.9 to 1.0. Precision in adult patients with respiratory insufficiency revealed excellent precision amongst measurements. 3. Electrical Impedance Tomography (EIT) Electrical impedance tomography capitalizes on changes in impendence in air-filled versus tissue-filled spaces to characterize and quantify regional distribution of lung volume at the bedside. Significant work has been done in the past decade to validate the technology in animals25 and in humans26, 27. The technology utilizes a series of 16 electrodes placed across the patient's chest (Figure 6). As small currents are passed between the electrodes, impedance is measured between and amongst the series. Through a complex interrogation and manipulation of these impedance values, a two-dimensional image is formed (Figure 7), and has been shown to correlate with clinical and radiographic changes in patients27. In ten mechanically ventilated adults with ARDS, end-expiratory lung volume as determined by nitrogen washout correlated well with end-expiratory lung impedance with an r2 of 0.95.26 The ability to estimate lung volume non-invasively and in real time may significantly improve outcomes in patients with lung injury. Specifically, the ability to determine a patient's ideal functional residual capacity and ventilate them towards that goal may improve oxygen delivery by maximizing pulmonary compliance and minimizing pulmonary vascular resistance. This study seeks to utilize varying levels of PEEP to alter end-expiratory volume, using EIT and other surrogate measures to confirm efficacy, measuring oxygenation and shunt fraction as the clinical end points. Should this study demonstrate the ability to effectively recruit lung and minimize shunt fractions using an aggressive PEEP strategy, further studies of clinical benefit from this will be warranted. 4. Exhaled Breath Condensate (as measure of lung health) There is a growing body of evidence regarding changes in airway lining fluid (ALF) pH in acute and chronic respiratory diseases that are characterized, at least in part, by inflammation. It has been demonstrated that the pH of ALF is low (acidic) in multiple pulmonary inflammatory diseases including asthma28, cystic fibrosis29, pneumonia, and ARDS30-32, and that this pH can be detected continuously, safely and non-invasively in exhaled breath condensate (EBC)33. The pH of EBC may be a safe, non-invasive screening tool for progression of ARDS, and of lung recruitment. It has anecdotally been shown to predict respiratory failure and impending respiratory infection (unpublished data). As seen in Figure 4 (left), the EBC pH is a marker exhibiting rapid turnover and thus may be valuable for real-time monitoring of lung pathology. Continuous exhaled breath condensate pH collection and assay system (ALFA monitor, Respiratory Research, Inc., Austin, Texas) consists of a condenser attached to the expiratory limb of the ventilator. Exhaled breath condensate is collected continuously from the expiratory port, condensed in a cooling chamber, CO2 removed, and collected in an inferior chamber where pH is continuously read. This yields a continuous, responsive measure from ventilated patients, which (1) takes samples from an exhaust port on the outside of the ventilator circuit, and (2) adds no measurable resistance to the ventilator circuit. The measurement of EBC in patients with lung injury may serve as an early marker of derecruitment. II. Study Objectives Specific Aim 1: To demonstrate the efficacy of the a maximal recruitment strategy to increase lung volumes and improve oxygenation in children with acute lung injury, utilizing multiple breath nitrogen washout (MBNW) and electrical impedance tomography (EIT) as measures of lung volume. (Hypothesis: Lung volumes and oxygenation will increase following the maximal recruitment protocol as compared to those during 'baseline ventilation' or the 'open lung approach.') Specific Aim 2: To compare lung volumes as measured by MBNW and EIT at varying end-expiratory lung volumes. (Hypothesis: Lung volumes as measured by MBNW will correlate with those obtained by EIT.)