Effects of positive end-expiratory pressure on gastric mucosal perfusion in acute respiratory distress syndrome

Introduction Positive end-expiratory pressure (PEEP) improves oxygenation and can prevent ventilator-induced lung injury in patients with acute respiratory distress syndrome (ARDS). Nevertheless, PEEP can also induce detrimental effects by its influence on the cardiovascular system. The purpose of this study was to assess the effects of PEEP on gastric mucosal perfusion while applying a protective ventilatory strategy in patients with ARDS. Methods Eight patients were included. A pressure–volume curve was traced and ideal PEEP, defined as lower inflection point + 2 cmH2O, was determined. Gastric tonometry was measured continuously (Tonocap). After baseline measurements, 10, 15 and 20 cmH2O PEEP and ideal PEEP were applied for 30 min each. By the end of each period, hemodynamic, CO2 gap (gastric minus arterial partial pressures), and ventilatory measurements were performed. Results PEEP had no effect on CO2 gap (median [range], baseline: 19 [2–30] mmHg; PEEP 10: 19 [0–40] mmHg; PEEP 15: 18 [0–39] mmHg; PEEP 20: 17 [4–39] mmHg; ideal PEEP: 19 [9–39] mmHg; P = 0.18). Cardiac index also remained unchanged (baseline: 4.6 [2.5–6.3] l min-1 m-2; PEEP 10: 4.5 [2.5–6.9] l min-1 m-2; PEEP 15: 4.3 [2–6.8] l min-1 m-2; PEEP 20: 4.7 [2.4–6.2] l min-1 m-2; ideal PEEP: 5.1 [2.1–6.3] l min-1 m-2; P = 0.08). One patient did not complete the protocol because of hypotension. Conclusion PEEP of 10–20 cmH2O does not affect gastric mucosal perfusion and is hemodynamically well tolerated in most patients with ARDS, including those receiving adrenergic drugs.


Introduction
Recent studies have shown that lung protective strategies using low tidal volumes and high levels of positive end-expiratory pressure (PEEP) reduce mortality and are becoming standard practice in patients with acute respiratory distress syndrome (ARDS) [1,2].
Although PEEP improves arterial oxygenation, it can adversely affect systemic hemodynamics, reducing venous return and cardiac output. These effects are proportional to the PEEP level. Regional perfusion can also be affected by PEEP, inde-pendently of cardiac output changes. The splanchnic perfusion is particularly sensitive, and any reduction can compromise its barrier function, promote bacterial translocation, and contribute to the development of multiple organ failure [3]. In experimental models, PEEP has markedly decreased mesenteric and portal blood flow, despite only moderate reductions in cardiac output [4][5][6][7][8]. Similar results have been reported in patients without lung injury [9,10]. These effects are usually dose related, becoming more pronounced with PEEP levels around 20 cmH 2 O. R307 Kiefer reported that PEEP did not significantly alter splanchnic blood flow in six patients with acute lung injury [11]. Nevertheless, caution should be exercised in extending these results to clinical practice, because only hemodynamically stable patients without adrenergic drugs were studied, and PEEP levels never exceeded 14 cmH 2 O [12].
Our aim was to evaluate the effects of PEEP levels up to 20 cmH 2 O on gastric mucosal perfusion and systemic hemodynamics in mechanically ventilated patients with ARDS under hemodynamic support.

Patients
The study was approved by the Ethics Committee of the Medicine Faculty and was performed in the Surgical Intensive Care Unit of the Catholic University Hospital of Chile.
Adult mechanically ventilated patients were considered eligible for the study if they met the following criteria for ARDS during the 24 hours that preceded the study: acute onset of respiratory failure; diffuse bilateral infiltrates in the chest radiograph; a ratio of partial pressure of O 2 (PaO 2 ) to fraction of inspired oxygen (FiO 2 ) of less than 200 mmHg; and a pulmonary arterial occlusion pressure less than 18 mmHg and no cardiac failure.
Hemodynamic monitoring included an arterial line and a pulmonary artery catheter (Baxter Edwards Critical-Care, Irvine, CA). Patients could be under vasopressor or inotropic support, but had to be hemodynamically stable for at least 3 hours before starting the protocol.
Patients were excluded if they had any of the following conditions: pregnancy, pre-existing respiratory dysfunction, cardiac index of less than 2.5 l min -1 m -2 , or were receiving enteral nutrition.

Interventions
A nasogastric tonometer (TRIP ® Tonometry Catheter 14F, with biofilter connector for TONOCAP™ Monitor) was inserted into the stomach and connected to air automated tonometry (TONOCAP™ Monitor; Datex-Engstrom, Helsinki, Finland). All patients were sedated with midazolam and morphine, and paralyzed with rocuronium. Neuromuscular relaxation was measured by a TOF watch ® device. An intravenous 20 mg dose of famotidine was administered before starting the protocol. Patients were connected to volume-controlled mechanical ventilation (Servo 900 C; Siemens, Solna, Sweden). A pressure-volume curve was obtained for each patient by the airway occlusion technique [13], and ideal PEEP was defined as the lower inflection point + 2 cmH 2 O, or 12 cmH 2 O if no lower inflection point was found. PEEP levels of 10, 15, 20 cmH 2 O, and ideal PEEP, with tidal volumes of 8 ml kg -1 , were applied in four consecutive 30 min periods, respectively. Respiratory rate was modified to maintain end tidal CO 2 within ± 10 mmHg of basal. All patients were receiving a constant infusion of 6% hetastarch before the beginning of the study (40-80 ml h -1 ). Cardiac output was optimized before and during the trial by determining the respiratory variation of systolic arterial pressure [14]. Whenever the variation was more than 10% a 100 ml bolus of 6% hetastarch was infused and the volume status was reassessed. No change in adrenergic support was allowed during the protocol. If hypotension (mean arterial pressure < 65 mmHg) persisted for more than 1 min, the protocol was stopped.

Measurements
At baseline, and at the end of each period, hemodynamic, ventilatory and tonometric measurements were performed, and arterial blood samples withdrawn. Hemodynamic records included mean arterial pressure, heart rate, cardiac output, pulmonary artery occlusion pressure, central venous pressure and left ventricular stroke work index. Cardiac output was measured by thermodilution as the average of three values obtained after injections of 10 ml of 5% dextrose in water at room temperature. Mean airway pressure, oxygenation index and PEEP levels were registered. Oxygenation index was calculated as mean airway pressure × FiO 2 × 100/PaO 2 . The CO 2 gap (gastric partial pressure of CO 2 [pCO 2 ] minus arterial pCO 2 ) was calculated by comparing simultaneous measurements of tonometric gastric mucosal pCO 2 and arterial pCO 2 .

Statistical analysis
Results are presented as median and range. The software Statview 5.0 was used to perform the statistical analysis. Nonparametric tests were used because of the small sample size. Data were analyzed with a Friedman test followed by a Wilcoxon signed-rank test if necessary. Results were considered statistically significant at P < 0.05.

Results
Eight patients with ARDS were enrolled. They had a median (range) age of 63.5 years (23-86), and an Acute Physiology and Chronic Health Evaluation II score of 14 (7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20) at admission to the intensive care unit. On the day of the study they had a median Sepsis-related Organ Failure Assessment (SOFA) [15] score of 10 (7-13). All patients fulfilled criteria for ARDS, as defined by the inclusion criteria, during the 24 hours before the study and they had been on mechanical ventilation for 32 (12-72) hours. They were being ventilated with a median PEEP level of 9 (4-12) cmH 2 O, they had a PaO 2 /FiO 2 ratio of 235 (144-388) mmHg and their respiratory system compliance was 45 (27-60) ml per cmH 2 O. Seven patients had sepsis (two pneumonia and five extrapulmonary sepsis), and one a severe thoracic trauma. Of the septic patients, six were in septic shock. Characteristics of individual patients are shown in Table 1.
No changes in cardiac index or in CO 2 gap were found at any of the study periods (Table 2). Oxygenation index, mean arterial pressure, pulmonary mean arterial pressure, pulmonary artery occlusion pressure, central venous pressure and left ventricular stroke work index also remained stable through the study. Only mean airway pressure and PaO 2 /FiO 2 ratio differed between periods, as expected. Five patients required a 100 ml bolus of hetastarch during the trial; in no patient was it necessary to repeat it. Individual changes in CO 2 gap and cardiac index are presented in Figs 1 and 2, respectively. At baseline three patients had already a CO 2 gap of more than 20 mmHg. After starting the protocol with 10 cmH 2 O PEEP, patient 6, who was previously being ventilated with 4 cmH 2 O PEEP, had a further increase in CO 2 gap. When PEEP was increased from 10 to 15 cmH 2 O, six patients decreased their CO 2 gap and two increased it. When PEEP was increased from 15 to 20 cmH 2 O, three patients increased their CO 2 gap, three decreased it and in one patient it remained unchanged. Patient 4 did not complete the protocol because of moderate hypotension (mean arterial pressure 60 mmHg) when PEEP was increased to 20 cmH 2 O. This patient recovered after an increased dose of norepinephrine (noradrenaline) and a return of PEEP to baseline levels.
Six of the eight patients studied survived (75%). The median length of stay in the intensive care unit was 17 (10-34) days and the median duration of mechanical ventilation was 9 (5-34) days.

Discussion
Our results show that high PEEP levels (up to 20 cmH 2 O) do not compromise gastric mucosal perfusion, as assessed by tonometry, and do not affect systemic hemodynamics in most patients with ARDS. This is consistent with the findings of two other studies on the effects of PEEP on splanchnic perfusion in patients with ARDS. Nevertheless, in contrast with our study, neither of those studies included patients in septic shock or under adrenergic support [11,16].
Shock and cardiovascular dysfunction are frequently associated with ARDS. This is an important issue, because hemodynamic safety concerns could preclude the use of high or optimal PEEP levels in that setting, even if necessary. A major finding of our study is that PEEP levels up to 20 cmH 2 O can be well tolerated, even in patients with ARDS and septic shock. Nevertheless, our trial was relatively short and we cannot exclude the possibility that keeping high PEEP levels for a longer period might result in increased fluid requirements, which could be deleterious in the longer term.
Experimental and clinical research has demonstrated that in mechanically ventilated subjects without lung injury, PEEP decreases venous return and, secondarily, cardiac output [17][18][19]. In addition, Trager and colleagues showed that, in patients with acute respiratory failure associated with septic shock, high PEEP levels induced a decrease in cardiac output [20]. In contrast, we found no decrease in cardiac output in our patients tested with increasing PEEP levels when fluid administration was optimized according to respiratory variation in systolic arterial pressure. A similar result was reported by Kiefer and colleagues and by Akinci and colleagues  [11,16]. Possible explanations for these contradictory results are a higher rate of fluid administration and the use of lower tidal volumes in the latter studies. Although we did not determine the upper inflection point of the pressure-volume curve, we think that by keeping tidal volume at 8 ml kg -1 any overdistension of the lungs was minimized. Lung volumes are a critical component of the hemodynamic effects of ventilation [21]. Thus, it seems that it is possible to preserve cardiac output in patients with ARDS, despite the use of high PEEP levels, by optimizing fluid administration and limiting tidal volumes.
Gastric mucosal perfusion, as assessed by CO 2 gap, also remained unchanged during the PEEP trial. This is consistent with the results reported by Kiefer and Akinci in similar studies. In all these studies cardiac output remained unchanged [11,16]. In contrast, Trager reported, in a series of septic shock patients with acute respiratory failure, that an increase in PEEP from 5 to 15 cmH 2 O induced a decrease in cardiac output associated to a decrease in hepatic vein O 2 saturation and in hepatic glucose production [20]. It therefore seems that by avoiding decreases in cardiac output, splanchnic perfusion can be preserved in the majority of the patients.
In spite of the fact that no significant changes in CO 2 gap or cardiac index were found during the protocol, when looking at the individual data certain patients evidenced an adverse effect when their PEEP level was increased. Patient 4, who had an extrapulmonary ARDS, presented hypotension when 20 cmH 2 O PEEP was applied. In this case, no simultaneous records of cardiac output or CO 2 gap could be made for safety reasons (we immediately proceeded to decrease PEEP level).

Figure 1
Individual changes in CO 2 gap (gastric pCO 2 minus arterial pCO 2 ) with different positive end-expiratory pressure levels Individual changes in CO 2 gap (gastric pCO 2 minus arterial pCO 2 ) with different positive end-expiratory pressure levels.

Figure 2
Individual changes in cardiac index with different positive end-expiratory pressure levels Individual changes in cardiac index with different positive end-expiratory pressure levels. One major limitation of our study is the small number of patients studied. Thus, a type II error cannot be excluded. We did not perform any a priori power analysis because we had no estimation of the possible magnitude of the effects that PEEP could have on gastric tonometry.
Another limitation is the rather moderate severity of ARDS in our study. Although all patients fulfilled criteria for ARDS during the 24 hours that preceded the study, at inclusion their PaO 2 /FiO 2 ratio and their respiratory system compliance were only moderately decreased. Two recent papers provide an explanation for this observation [22,23]. They show in patients diagnosed with ARDS that after a few hours of treatment with PEEP or a high FiO 2 , more than half of the patients present a PaO 2 /FiO 2 ratio of more than 200 mmHg. In addition, respiratory system compliance increased by more than 10 ml per cmH 2 O after 6 hours of treatment with PEEP [23]. At inclusion our patients had already been ventilated with a median PEEP level of 9 cmH 2 O for more than 12 hours, which could have explained the rather improved respiratory performance at baseline. In any event, this improvement demonstrated a less severe ARDS. It is possible that more severely compromised patients might present a lower tolerance to high PEEP levels.
Other limitation is that tonometry was the sole method used to assess gastric mucosal perfusion. Nevertheless, Elizalde and colleagues showed that gastric mucosal blood flow, measured by laser Doppler flowmetry and by reflectance spectrophotometry, is well correlated with gastric intramucosal acidosis in mechanically ventilated patients [24].

Conclusions
Our study supports the findings of previous studies suggesting that high PEEP levels do not affect splanchnic perfusion and are hemodynamically well tolerated in most patients with ARDS. Furthermore, our study shows that gastric mucosal perfusion can be well preserved while high PEEP levels are applied even in patients presenting cardiovascular dysfunction and receiving adrenergic support, which is a frequent occurrence in critical care. Nevertheless, two of the eight patients studied exhibited adverse effects during the PEEP trial, which highlights the importance of a close monitoring of systemic and regional perfusion while applying high PEEP levels to patients with ARDS. Future studies should assess the effects of PEEP on splanchnic perfusion in a longer term.