Changes of splanchnic perfusion after applying positive end expiratory pressure in patients with acute respiratory distress syndrome.

BACKGROUND: 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.
MATERIALS AND METHODS: Thirty-two patients were included in the study. A pressure-volume curve was traced and ideal PEEP, defined as lower inflection point + 2cmH(2)O, was determined. Gastric tonometry was measured continuously (Tonocap). After baseline measurements, 10, 15 and 20cmH(2)O PEEP and ideal PEEP were applied for 30 min each. By the end of each period, hemodynamics, CO(2) gap (gastric minus arterial partial pressures), and ventilatory measurements were taken.
RESULTS: PEEP had no effect on CO(2) gap (median [range], baseline: 18 [2-30] mmHg; PEEP 10: 18 [0-40] mmHg; PEEP 15: 17 [0-39] mmHg; PEEP 20: 16 [4-39] mmHg; ideal PEEP: 19 [9-39] mmHg; P = 0.19). Cardiac index also remained unchanged (baseline: 4.7 [2.6-6.2] l min(-1) m(-2) ; PEEP 10: 4.4 [2.5-7] l min(-1) m(-2) ; PEEP 15: 4.4 [2.2-6.8] l min(-1) m(-2) ; PEEP 20: 4.8 [2.4-6.3] l min(-1) m(-2) ; ideal PEEP: 4.9 [2.4-6.3] l min(-1) m(-2) ; P = 0.09).
CONCLUSION: PEEP of 10-20 cmH(2)O does not affect splanchnic perfusion and is hemodynamically well tolerated in most patients with ARDS, including those receiving inotropic supports.

Introduction

Many recent studies have shown that mechanical ventilation (MV) in patients with acute respiratory distress syndrome (ARDS), using low tidal volumes and high levels of positive end-expiratory pressure (PEEP) reduces the mortality rate and this ventilatory strategy is now accepted as standard practice in patients with ARDS.[12]

Although PEEP improves arterial oxygenation, it has adverse hemodynamic effects. PEEP reduces the venous return to the heart and the left ventricular end diastolic volume by pushing the inter ventricular septum towards the left, and thus reduces the cardiac output. These effects are proportional to the PEEP level. Regional perfusion can also be affected by PEEP, independently 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.[4-8] Similar results have been reported in patients without lung injury.[910] These effects are usually dose related, becoming more pronounced with PEEP levels around 20 cmH2O.

Kiefer reported that PEEP did not significantly alter splanchnic circulation 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 cmH2O.[12]

Our aim was to evaluate the effects of PEEP levels up to 20 cmH2O on gastric mucosal perfusion and systemic hemodynamics in mechanically ventilated patients with ARDS on hemodynamic support.

Materials and Methods

Patients

The study was performed in the Intensive Care Unit (department of Anesthesiology) of the Banaras Hindu University Hospital, Varanasi, India.

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 involving more than three-fourths of both lung fields; a ratio of partial pressure of O2 (PaO2) to fraction of inspired oxygen (FiO2) 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. Patients could be on vasopressor or inotropic support (Dopamine 5-20 microgram/kg/min with or without Noradrenaline 10-20 micrograms /min dose), but had to be hemodynamically stable (with a mean arterial blood pressure of more than 60mm of Hg, pulse rate should be less than 100/min and more than 50 per minute) for at least 5 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 an air automated tonometer (TONOCAP™ Monitor; Datex-Engstrom, Helsinki, Finland). All patients were sedated with midazolam, and paralyzed with vecuronium. Neuromuscular relaxation was measured by a Train of Four watch® device. A 40mg intravenous dose of pantoprazole was administered before starting the study. Patients were started on volume controlled ventilation. 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 cmH2O, or 12 cmH2O if no lower inflection point was found.

PEEP levels of 10, 15, 20 cmH2O, and ideal PEEP, with tidal volumes of 6 ml kg−1, were applied in four consecutive 30 min periods, respectively. Respiratory rate was modified to maintain end tidal CO2 within ±10mmHg of basal. All patients were receiving a constant infusion of 6% hetastarch before the beginning of the study (40–60 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 changes in vasopressor or inotropic support were allowed during the study. If hypotension (mean arterial pressure <60 mmHg) persisted for more than 2 min, the protocol was stopped.

Measurements

Hemodynamic, ventilatory and tonometric measurements were made at baseline, and at the end of each period, and arterial blood samples taken. 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 10ml 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 × FiO2 × 100/PaO2. The CO2gap (gastric partial pressure of CO2[pCO2] minus arterial pCO2) was calculated by comparing simultaneous measurements of tonometric gastric mucosal pCO2 and arterial pCO2.

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 was analyzed with a Friedman test followed by a Wilcoxon signed-rank test if necessary. Results were considered statistically significant at P < 0.05.

Results

Thirty-two patients with ARDS were enrolled. They had a median (range) age of 47 years, (25 were male and seven were female, and two of the female patients were in postpartum state, four patients had previous history of diabetes) and an Acute Physiology and Chronic Health Evaluation II score of 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. 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) cmH2 O, had a PaO2/FiO2 ratio of 230 (140–386) mmHg and their respiratory system compliance was 45 (27–60)ml per cmH2O. Twenty-eight patients had sepsis (eight pneumonia and 20 extra pulmonary sepsis), and four had severe head and thoracic injury. Of the septic patients, 24 were in septic shock.

No changes in cardiac index or in CO2 gap were found at any of the study periods [Table 1]. 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 PaO2/FiO2 ratio differed between periods, as expected. Twenty patients required a 100ml bolus of hetastarch during the trial; in no patient was it necessary to repeat it. At baseline, 12 patients had already a CO2 gap of more than 20 mmHg. After starting the protocol with 10 cmH2O PEEP, 24 patients decreased their CO2 gap and eight increased it. When PEEP was increased from 15 to 20 cmH2O, 12 patients increased their CO2 gap, 12 decreased it and in four patients it remained unchanged.

Table 1

Respiratory, hemodynamic and tonometric measurements

Parameter	Baseline (n = 32)	PEEP 10 (n = 32)	PEEP 15 (n = 32)	PEEP 20 (n = 32)	Ideal PEEP (n = 32)	P
PEEP (cmH2O)	9 (4–12)	10	15	20	12 (8–15)
Mean airway pressure (cmH2O)	12.1 (8–18.7)	14.3 (12–17)	20.2 (17–22.2)	24.7 (22–26.4)	15.9 (11.5–22.2)	0.0001a
OI (cmH2O per mmHg)	5.3 (2.9–12.4)	7 (3–14.5)	6.7 (4.1–12.3)	7 (5–12.3)	6.6 (2.9–12.3)	0.31
PaO2/FiO2 (mmHg)	230 (140–386)	208 (115–412)	281 (151–421)	329 (197–438)	240 (169–469)	0.0009b
PaCO2 (mmHg)	34 (31–51)	42 (28–61)	43 (31–66)	46 (36–59)	43 (29–52)	0.09
Cardiac index (I min−1 m−2)	4.7 (2.6–6.2)	4.4 (2.5–7)	4.4 (2.2–6.8)	4.8 (2.7–6.1)	4.9(2.4–6.3)	0.09
LVSWI (g mm−2)	45 (22–71)	43 (22–60)	40 (14–60)	36 (15–58)	42 (14–66)	0.13
MAP (mmHg)	76 (74–103)	79 (69–99)	77 (72–97)	76 (67–93)	75 (71–96)	0.24
PAOP (mmHg)	15 (10–18)	16 (8–20)	16 (11–21)	18 (12–24)	15 (13–23)	0.23
CVP (mmHg)	10 (9–17)	12 (10–19)	13 (11–24)	12 (10–19)	13 (8–18)	0.25
CO2 gap (mmHg)	18 (2–30)	18 (0–40)	17 (0–39)	16 (4–39)	19 (9–39)	0.19

Results are presented as median (range). CVP-central venous pressure; CO2 gap-arterial partial pressure of CO2[pCO2] minus gastric pCO2; FiO2-fraction of inspired oxygen; LVSWI-left ventricular stroke work index; MAP-mean arterial pressure; OI-oxygenation index-defined as mean airway pressure × FiO2 × 100/arterial pCO2; PaO2-partial pressure of O2; PaCO2-partial pressure of CO2; PAOP-pulmonary arterial occlusion pressure; PEEP-positive end-expiratory pressure.

P < 0.05 for all comparisons except baseline versus PEEP 10 and PEEP 10 versus ideal PEEP.

P < 0.05 for all comparisons except baseline versus PEEP 10, baseline versus PEEP 15, baseline versus ideal PEEP, and PEEP 15 versus ideal PEEP

Twenty-four of the 32 patients studied, survived (75%). The median length of stay in the intensive care unit was 18 days and the median duration of mechanical ventilation was 11 (5–34) days.

Discussion

Similar studies have been done previously, but the number of cases studied were very few; we have done this study with a relatively large number of patients and our results show that high PEEP levels (up to 20 cmH2O) 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. However, in contrast to our study, neither of those studies included patients in septic shock or on adrenergic support.[1116]

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 cmH2O 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.[17-19] 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 the respiratory variation in systolic arterial pressure. A similar result was reported by Kiefer et al. and by Akinci et al.[1116] 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 6 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 CO2 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.[1116] In contrast, Trager reported, in a series of septic shock patients with acute respiratory failure, that an increase in PEEP from 5 to 15 cmH2O induced a decrease in cardiac output associated with a decrease in hepatic vein O2 saturation and in hepatic glucose production.[20] It therefore seems that by avoiding a decrease in cardiac output, splanchnic perfusion can be preserved in a majority of patients.

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 the criteria for ARDS during the 24 hours that preceded the study, at inclusion their PaO2/FiO2 ratio and their respiratory system compliance were only moderately decreased. Two recent papers provide an explanation for this observation.[2223] They show in patients diagnosed with ARDS that after a few hours of treatment with PEEP or a high FiO2, more than half of the patients present a PaO2/FiO2 ratio of more than 200 mmHg. In addition, the respiratory system compliance increased by more than 10 ml per cmH2O after 6 hours of treatment with PEEP.[23] At inclusion our patients had already been ventilated with a median PEEP level of 9 cmH2O 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.

Another limitation is that tonometry was the sole method used to assess gastric mucosal perfusion. Nevertheless, Elizalde et al. 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 vasopressor support, which is a frequent occurrence in critical care. Future studies should assess the effects of PEEP on splanchnic perfusion in a longer term.