Preliminary study of ventilation with 4 ml/kg tidal volume in acute respiratory distress syndrome: feasibility and effects on cyclic recruitment - derecruitment and hyperinflation

The main mechanisms of ventilator-induced lung injury (VILI) are overdistension and cyclic recruitment-derecruitment (R/D). While low tidal volume (Vt) has been applied successfully with the rationale of preventing overdistension [1], high positive end-expiratory pressure (PEEP) strategies, which should theoretically help to prevent cyclic R/D, have not been able to show a clear benefit on survival in acute respiratory distress (ARDS) patients [2–4]. One of the main drawbacks of increasing PEEP to prevent cyclic R/D is that a parallel increase in stress and strain occurs at end-inspiration.

We recently showed that Vt is also an important determinant of cyclic R/D, and that high Vt may induce injury not only by inducing overdistension, but also cyclic R/D. However, we observed that at a Vt of 6 ml/kg there was still evidence of cyclic R/D [5]. Other studies have also shown that a Vt of 6 ml/kg may still be associated to hyperinflation, which has been shown to be a relevant marker of VILI [6, 7]. Therefore, it has been proposed that in order to achieve higher levels of protection, it may be necessary to move toward ultra-protective strategies [6, 8, 9]. High-frequency oscillatory ventilation or extracorporeal lung support techniques have been suggested as potential alternatives in this direction. However, these interventions require rather complex technologies, not widely available, and which have still not proven to be superior to conventional protective mechanical ventilation.

An alternative strategy may be to optimize conventional protective mechanical ventilation by further decreasing Vt below 6 ml/kg. Based on our previous findings, we hypothesize that this intervention may further decrease cyclic R/D, but in contrast to increasing PEEP, a lower Vt may also decrease the risk of overdistension. However, an important concern with using very low Vt is the risk of severe hypercapnia and respiratory acidosis.

The goal of this preliminary physiologic study was to determine if decreasing Vt from 6 to 4 ml/kg, with conventional mechanical ventilation, is effective to reduce cyclic R/D and hyperinflation, and if it is possible to prevent severe hypercapnia and respiratory acidosis at such very low Vt, without using extracorporeal lung support techniques.

We designed a crossover study at two Vt levels: 6 and 4 ml/kg predicted body weight (PBW). The protocol had two parts: one at the bedside and other at the computed tomography (CT) room. The Research Ethics Committee of the School of Medicine of the Universidad Católica de Chile granted approval for this study, and informed consent was obtained from the patients' next of kin. We included patients with acute onset of hypoxemia (PaO2/FiO2 <300), bilateral opacities in chest X-ray, without evidence of left ventricle overload, connected to mechanical ventilation for less than 48 hours, and who required a lung CT scan according to the treating physician.

In this preliminary physiologic study we showed in 10 ARDS patients that by decreasing Vt from 6 to 4 ml/kg we were able to reduce cyclic R/D, as well as hyperinflated tissue, plateau pressures and driving pressures. Severe respiratory acidosis and severe hypercapnia could be effectively prevented at Vt 4 ml/kg by increasing respiratory rate and reducing instrumental dead space.

Experimental studies using dynamic CT, subpleural videomicroscopy and PaO₂ oscillations, have shown evidence that cyclic R/D may occur in ARDS [17–19]. Recently, by applying dynamic CT during mechanical ventilation we provided evidence that cyclic R/D also occurs in ARDS patients [5]. In that study, we demonstrated that decreasing Vt from a high to a low Vt (12 to 6 ml/kg) markedly decreased cyclic R/D. Now, we showed that cyclic R/D may be further decreased by applying a very low Vt, such as 4 ml/kg. Although a 0.7% absolute reduction in R/D may appear as modest, CT scan may systematically underestimate R/D because of the limited resolution of CT images, which defines that the mean CT attenuation measured on each voxel is an average of different alveolar densities. Therefore, voxels showing apparently aerated units at end expiration may contain several collapsed and unstable alveoli mixed with aerated units. In fact, videomicroscopic images obtained in experimental models confirm that unstable alveoli can be adjacent to stable alveoli.

Besides the potential benefit of reducing cyclic R/D, reduction in Vt decreased airway pressures, specifically plateau pressure. Hyperinflated tissue decreased in parallel to plateau pressure. Therefore, these data suggest that a very low Vt may prevent both, overdistension and cyclic R/D. Recently, Terragni et al. showed in ARDS patients that decreasing Vt from 6.3 ± 0.2 to 4.2 ± 0.3 ml/kg, was associated to a decrease in tidal hyperinflation, which was paralleled by a decrease in pulmonary inflammation, as assessed by cytokines in bronchoalveolar lavage [6]. However, in their study, Vt reduction was associated to an increase in PEEP levels, and hypercapnia was avoided by the use of extracorporeal CO₂ removal.

One of the major concerns of using a very low Vt is the risk of hypercapnia. Richecoeur et al. showed several years ago that rather simple interventions such as increasing respiratory rate and decreasing instrumental dead space could be effective to control hypercapnia in ARDS patients [20]. Therefore, during ventilation with 4 ml/kg we increased respiratory rate to keep constant minute ventilation. In addition, we reduced instrumental dead space by replacing the HMEF by a heated humidifier. This intervention has been shown to be effective to reduce PaCO₂ while ventilating ARDS patients with a low Vt strategy [11, 21]. Although some authors have suggested that HMEFs should be avoided during ventilation with a Vt of 6 ml/kg [22], this is not a standard practice, and furthermore, several ICUs keep using HMEFs in mechanically ventilated ARDS patients [23, 24]. In the present study we kept the HMEF during ventilation with Vt 6, as was the usual practice in our ICU, but we removed it during ventilation with Vt 4. Although this variation in dead space between Vt 6 and Vt 4 limits the possibility of comparing PaCO₂ and pH with both Vts, the purpose of this part of the study was just to determine if severe hypercapnia and respiratory acidosis could be avoided during ventilation with Vt 4. Although 30 minutes may seem a short period, it has been shown that it is enough to reach a new equilibrium in PaCO₂ after changes in ventilation [25, 26]. Our data suggest that a ventilatory strategy based on a Vt of 4 ml/kg would be feasible in several ARDS patients. However, we acknowledge that only one patient had a severe ARDS, and therefore, future studies should determine whether this intervention can be applied in severe ARDS patients without inducing severe respiratory acidosis or requiring CO₂ extracorporeal removal.

In the present study we used respiratory rates as high as 39/minute during ventilation with Vt 4. Interestingly, we did not observe auto-PEEP in any patient, probably because expiratory time decreased in the same proportion as Vt and ARDS patients typically have very short time constants. A second concern is that higher respiratory rates may favor VILI, as has been shown in experimental data [27, 28]. In fact, although hyperinflation may decrease in magnitude because of the lower Vt, peak inspiratory hyperinflation phenomena occur more frequently because of the compensatory increase in respiratory rate, which could eventually counteract the potential benefit of decreasing Vt. However, the results of the ARDS network tidal volume study suggest that the potential disadvantage of a larger respiratory rate may be largely surpassed by the benefits of a lower Vt [1]. The use of extracorporeal lung support may allow both Vt and respiratory rate to decrease, although it has its own risks.

Another potential disadvantage of further decreasing Vt is a progressive alveolar derecruitment secondary to lower mean and plateau airway pressures. This derecruitment may be prevented by a parallel increase in PEEP levels [6, 29, 30]. However, as we were interested in the isolated impact of Vt on cyclic R/D, we did not modify PEEP levels. Despite keeping PEEP constant and lowering airway pressures, we did not observe an increase in non-aerated tissue in CT. However, CT images were obtained after only 5 minutes of ventilation. It is possible that after a longer period derecruitment could have been observed [29]. Nevertheless, PaO₂/FiO₂ did not change after 30 minutes of ventilation with 4 ml/kg.

In this preliminary physiologic study we observed that decreasing Vt from 6 to 4 ml/kg reduces cyclic R/D, hyperinflation, as well as airway pressures. In addition, we showed that by optimizing instrumental dead space and respiratory rate, it is possible to avoid severe hypercapnia and respiratory acidosis, while ventilating ARDS patients with a Vt of 4 ml/kg PBW. However, these findings must be confirmed in larger studies and complemented with assessment of relevant biomarkers of lung injury, before it can be translated to clinical practice.