Introduction

In acute respiratory failure, improved survival was demonstrated with low tidal volume (VT) ventilation [1]. Consequently, the use of small VT has been recommended during mechanical ventilation [2], but this strategy may facilitate alveolar de-recruitment and deterioration of gas exchange [3, 4]. Recruitment maneuver (RM) may improve gas exchange [3, 4, 5], but inflating the lungs to nearly vital capacity might be harmful due to stretch stress imposed on the pulmonary parenchyma. Alveolar macrophages liberate inflammatory cytokines in response to stretch [6], and injurious mechanical ventilation is associated with higher levels of inflammatory cytokines than protective ventilation [1, 7, 8]. Thus, inflammatory mediators released from the lung are proposed to contribute to ventilator-associated lung injury [9, 10]. The increase and decrease of systemic mediator levels occur rapidly if ventilation is changed [8], but the time period of stretch stress necessary to induce the production or liberation of mediators is unclear. In small animals hyperinflating the lungs for brief periods can alter microvascular permeability [11]. Thus, we hypothesized that one RM may potentially affect the balance of pro- and anti-inflammatory mediators and tested this hypothesis by studying systemic mediator levels in patients with radiological and clinical signs of atelectasis before and after a single RM. Inflating the lungs for 7–8 s only with an airway pressure (Paw) of 40 cmH2O has been shown to effectively recruit atelectasis [12]. This short inflation period may decrease hemodynamic side effects and the risk for barotrauma compared to RMs with an extended inflation period and was therefore chosen in the present study.

Material and methods

With approval of the local ethics committee (application 4/7/02), 16 mechanically ventilated patients were enrolled into the study. Informed consent to participate in the study was obtained from their nearest relatives.

Inclusion criteria were:

  • Age > 18 years

  • Mechanical ventilation

  • Arterial catheter inserted

  • PaO2/FiO2 < 300 mmHg or decrease in PaO2 > 50 mmHg within the last hour

  • Radiological evidence of atelectasis in the chest X-ray

  • Informed consent

Exclusion criteria were:

  • Age < 18 years

  • Elevated intracranial pressure > 25 mmHg

  • Hemodynamic instability

  • Bronchopulmonary fistula

  • Emphysema

Patients were treated according to standard ICU care. Ventilatory settings before inclusion into the study were pressure-limited ventilation (pressure-controlled or pressure-regulated volume-guaranteed ventilation) with VT between 5 ml/kg and 8 ml/kg (ideal body weight), an I:E ratio between 1:1 and 1:2, PEEP and FiO2 adjusted to obtain an SaO2 of 90–95%. VT above 8 ml/kg ideal body weight were only permitted in patients with intracerebral lesions if hypercapnia could not be controlled otherwise. Blood pressure and SO2 were displayed on a bedside monitor (Datex AS/3, Datex Division Instrumentarium, Helsinki, Finland). Arterial blood samples were analyzed with the ABL 725 (Radiometer Copenhagen, Denmark).

Protocol

During the study, patients were positioned supine; ventilatory settings remained unchanged and all manipulations were avoided. Hemodynamic data were continuously recorded. At baseline arterial blood was sampled for blood gas- and cytokine analysis. Then, Paw of 40 cmH2O was applied for 7 s [12]. Thereafter, mechanical ventilation was continued, and arterial blood was sampled 5 min, 30 min, 60 min, 180 min and 360 min after the RM.

Ventilatory measurements

Gas flow and Paw were measured with the integrated monitoring of our ventilators (EVITA IV, Dräger, Lübeck, Germany or Servo-i Maquet, Solna, Sweden).

Cytokine measurements

Arterial blood samples were centrifuged at 3,000 cycles/s for 3 min; serum was aspirated and stored at −80°C. Simultaneous quantification of interleukin (IL)-1β, IL-6, IL-8, IL-10, IL-12p70 and tumor necrosis factor (TNF)-α levels was performed by a Cytometric Bead Array (Human Inflammation Kit, BD Biosciences, Heidelberg, Germany).

Statistical analysis

Results are expressed as mean ± SD. Statistical analysis was performed using the software package STATISTICA (StatSoft, Tulsa, OK, USA). An immediate effect of the RM on hemodynamics, gas exchange or respiratory compliance was assessed with a Wilcoxon matched-pairs test, by comparing baseline values with data obtained during the RM or 5 min thereafter. After the RM, the time course of these parameters was analyzed with a Friedman analysis of variance (ANOVA), which was also used to analyze the plasma cytokine concentrations throughout the study period. To compare single data points, Scheffé post-hoc analysis was performed. If cytokine levels were below the detection threshold, calculations were performed using the detection threshold as numerical values. p < 0.05 was chosen as the significance level.

Results

Patients

Patient characteristics are shown in Table 1. Ten of 16 patients met the ARDS (adult respiratory distress syndrome) or ALI (acute lung injury) criteria [13]. Fifteen patients had elevated levels of C-reactive protein and nine patients had elevated leucocytes, indicating an ongoing inflammatory response prior to the RM. One patient (No. 11) developed bronchospasm during the RM, which required anti-obstructive therapy and a change of the respiratory settings. He was therefore excluded from the analysis. Thus, 15 patients completed the study protocol without adverse events.

Table 1 Patients' characteristics (P Plateau inspiratory plateau pressure, V T tidal volume, CRP C-reactive protein, WBC white blood count)
Full size table

Gas exchange, respiratory mechanics and hemodynamics

At baseline, oxygenation and respiratory compliance were markedly impaired (Table 2). Oxygenation improved slightly, immediately after the RM, (p = 0.02) and increased even further thereafter (p = 0.019; Fig. 1). PaCO2 decreased after the RM (p = 0.006) and was subsequently stable (Table 2). Compliance was unchanged throughout the study period. Heart rate did not change over time, whereas the RM caused a moderate drop of the mean arterial pressure (MAP) (6 mmHg, p = 0.025).

Fig. 1
figure 1

Oxygenation before and after the recruitment maneuver. The time course of oxygenation expressed as PaO2/FiO2 on the y-axis is shown before and after the recruitment maneuver (RM) for individual patients. The mean value is marked by a bold line. Note that oxygenation improved only slightly immediately after the RM, but continued to improve throughout the study period, although ventilatory settings were not changed compared to baseline (-5 min)

Full size image
Table 2 Hemodynamic and ventilatory variables. Data are presented as mean ± SD. (Crs compliance of the respiratory system, HR heart rate, MAP mean arterial pressure, n.m. not measured, RM recruitment maneuver)
Full size table

Systemic levels of pro- and anti-inflammatory mediator

Before the RM, mediator concentrations above the lower detection limit were found in all patients for IL-6 and IL-8, in ten of 15 patients for IL-10 and IL-12p70, in seven of 15 patients for IL-1β and five of 15 patients for TNF-α levels. No change of the plasma mediator levels was observed during the study period (Fig. 2), regardless of whether all patients were analyzed, or a subgroup of five patients with the highest respective cytokine concentrations before the RM (Table 3).

Fig. 2
figure 2

Plasma mediator levels before and after the recruitment maneuver. The different panels show the time course of the plasmatic mediator concentrations before and after the recruitment maneuver (RM) for individual patients. In each panel, the mean value is marked by a bold line and the lower detection limit is given by a horizontal line. The plasmatic mediator concentrations were not affected by the RM

Full size image
Table 3 Systemic mediator concentrations (IL interleukin, RM recruitment maneuver, TNF tumor necrosis factor, Upper 33% subgroup of five patients with the highest respective plasma mediator level)
Full size table

Discussion

Serum-concentrations of pro- and anti-inflammatory mediators were not influenced by elevating Paw once to 40 cmH2O for 7 s, regardless of whether all patients were analyzed or only a subgroup of five patients with the highest respective cytokine levels prior to the RM.

Mediator levels

The influence of mechanical ventilation on pro- and anti-inflammatory mediators is rather complex [14]. Injurious mechanical ventilation may increase systemic cytokine levels in ARDS patients, but pro- and anti-inflammatory cytokines are affected at the same time [8]. In lung-healthy patients, a potentially injurious ventilation had no effects on systemic cytokine levels, even though the performed operating procedures induced some kind of immune response [15]. The majority of our patients had been treated in the ICU for more than 48 h, showed signs of an ongoing immune response and fulfilled ARDS or ALI criteria [13]. Surprisingly, the plasmatic mediator levels in our study were lower than those previously reported in ARDS patients [7, 8]. Thus, it is possible that our patients were not vulnerable to ventilator-associated lung injury, similarly to lung-healthy patients [15]. However, according to the results of a recent study, all of our patients were at risk for ventilator-associated lung injury [16], so that this explanation seems rather unlikely.

Recruitment maneuver

Different RMs have been described in an attempt to reopen atelectasis with minimal side effects [17]. In lung-healthy patients, a minimum Paw of 40 cmH2O is necessary to effectively recruit atelectasis [18], and approximately the same Paw has been applied in ARDS patients for sustained inflation RMs [17]. However, the optimal duration of such a sustained inflation has never been investigated in ARDS patients. In experimental lung injury [19] and patients during general anaesthesia [12], recruitment follows an exponential time course characterized by a maximum recruitment effect during the beginning of inflation, followed by a progressive decline of additionally recruited atelectasis over time. The time constants for recruitment averaged 1–4 s [12, 19]. Consequently, the RM performed in the present study should reopen 80–90% of the recruitable lung tissue, so that a longer inflation period would have been most likely with only little further effect.

Effect of the recruitment maneuver on pulmonary gas exchange and hemodynamics

Oxygenation and carbon dioxide elimination improved after the RM, indicating recruitment. However, a marked inter-individual variability was observed for the oxygenation effect, which in addition was clinically unimportant in the majority of patients (Fig. 1). This finding is in line with previous results, since the effect of a RM on gas exchange may depend on the etiology of respiratory failure, the duration of mechanical ventilation and the lung volume history [17]. The blood pressure drop during the RM is a well known phenomenon [17]. Since we limited the inflation period to 7 s, the decrease of MAP was moderate and clinically unimportant in most patients.

Conclusion

A single RM, performed by inflating the lungs with an airway pressure of 40 cmH2O for 7 s has no measurable effect on the systemic concentration of pro- and anti-inflammatory mediators in ventilated patients with clinical and radiological evidence of atelectasis.