Soon after publication of the ARMA trial, the study raised intense debate in the medical community and was criticized by its study design, use of surrogate markers of lung injury, and selection of the control group. Many experts in the field argued that a single ventilator strategy with low TVs may not be beneficial or applicable to all patients with ARDS due to the heterogeneity of the underlying lung injury that characterizes this population of critically ill patients.
The ARMA study was criticized for comparing 2 extremes of TV (6 mL/kg PBW vs 12 mL/kg PBW) rather than a safer range of TV (8-10 mL/kg PBW) which was the routine practice at the time of the trial. In the controversial meta-analysis by Eichacker et al the authors examined the results of 5 RCTs on low TV ventilation in ARDS and questioned the validity of the ARMA trial.11 The authors argued that in the 2 RCTs that showed a survival benefit from a low TV, the traditional arm received excessively large TVs (≥10 mL/kg) that were not the standard of care (8-9 mL/kg). In contrast, the 3 RCTs that showed no survival benefit from a low TV strategy used TVs in the traditional arm that resulted in lower airway pressures (Ppl 28-32 cm H2O). Therefore, the mortality benefit from a low TV strategy may reflect an excess mortality in the traditional arm exposed to unconventionally high pulmonary pressures (Ppl 34-37 cm H2O) rather a benefit from low TVs. This analysis and the authors’ conclusions have been vigorously challenged and rebutted by others. However, even though it did not prove the benefit of a low TV approach over traditional practice, this meta-analysis did indicate that using a high TV and high Ppl had deleterious effects.
Two subsequent meta-analyses by 2 independent groups of investigators combined the mortality data from the same 5 RCTs examined by Eichacker et al and performed statistical analysis to account for the overall clinical heterogeneity in these studies.12,23 The results of the treatment effect on mortality were similar. Petrucci and Lacovelli found that even though the 28-days mortality was significantly reduced by lung-protective ventilation (RR (relative risk): 0.74, 95%CI: 0.61-0.88), the mortality was not significantly different between low and conventional TV if the control groups kept the Ppl less than or equal to 31 cm H2O (RR: 1.13, 95%CI: 0.88-1.45).12 Similarly, the meta-analysis by Moran et al found that the pooled estimate of treatment effect favored protective ventilation but was not statistical significant. However, the treatment effect on 28-days mortality was significant for ARDS patients receiving a TV less than 7.7 mL/kg PBW if the control group had a Ppl more than or equal to 30 cm H2O but not if the controls had Ppl less than 30 cm H2O.23 These results remained unchanged in 2 subsequent meta-analyses by Petrucci et al which included all RCTs available to date.13,14 The authors suggested that their results may involve variations in transpulmonary pressure in the individual patient. A large TV might induce lung damage when the transpulmonary pressure is high. On the other hand, when transpulmonary pressures are within the safe range, a TV in the middle range (8-10 mL/kg) could be used avoiding deleterious effects. The results of this meta-analysis supports the belief of other leaders in the field that the TV in ARDS patients should be adjusted based on other markers of lung injury such as airway pressures or lung strain. Their rationale is that although some patients, such as those with poor pulmonary compliance and high airway pressures, would benefit from very low TV, others with less severe lung injury may require larger volumes to maintain ventilation and avoid alveolar collapse.
To proof this point, Deans et al provided data to support that the mortality in the ARMA trial depended on the lung compliance prior to randomization into a mechanical ventilation strategy.15 In patients with less compliant lungs (< 0.6 mL/cm H2O/kg PBW), there was a direct linear relationship with TV and these patients received lower TVs. There was no association in patients with better lung compliance (≥ 0.6 mL/cm H2O/kg PBW) and in those patients the TV remained constant. Furthermore, the effect on mortality by changing TV was significantly associated with the prerandomization lung compliance. In patients with more compliant lungs, a lower TV was associated with higher mortality. Conversely, in those patients with less compliant lungs, a low TV was associated with lower mortality compared to a high TV. These associations persisted even after accounting for differences in age, APACHE II, and PaO2/FiO2 ratio. Overall, these findings strongly suggest that mechanical ventilation should be managed according to other parameters than TV and Ppl and that a single ventilator strategy with a TV of 6 cc/kg PBW may be not be applicable to all patients with ARDS. To further support their findings, Deans et al also performed a retrospective analysis of 2587 patients screened by the ARDS Network who met enrollment criteria but were ineligible due to technical reasons (eg, difficulties with consent). This group of patients received routine care during the course of the ARMA trial and reflected the standard practice of the time. Remarkably, these ineligible patients for the ARMA trial had a comparable mortality rate to the low TV arm in that study (31.7% and 31%, respectively).
In a recent counterpoint analysis, Gattinoni challenged the use of low TV set at 6 mL/kg PWB because PBW is not an appropriate surrogate for the “resting lung volume” when defining lung strain as the ratio of TV to the resting lung volume.16 In patients with ARDS, PBW cannot be considered an acceptable surrogate for lung volume because the normal relationship between PBW, lung volume, and height is lost. Even though lung strain always increases with TV, the same applied TV/PBW may lead to completely different strain depending on the available lung volume still open to ventilation (“baby lung” volume) confirming that one size of TV may not fit all patients. In a subgroup of patients with very low “baby lung,” a TV 6 cc/kg PBW could be excessively high. In another subgroup of patients with a greater “baby lung,” a TV 6 cc/kg PBW could be unnecessarily low increasing the risk of atelectasis, respiratory acidosis, and need for supplementary sedation. This controversy stems from the difficulties of measuring transpulmonary pressure at the bedside—the real distending force of the lung and the cause of alveolar trauma. The ideal ventilation would measure the lung volume and transpulmonary pressure because IBW and Ppl are inadequate surrogates for lung stress and strain. However, it may not be feasible to have these measurements available on a routine basis in most ICUs and a lower TV/PBW might be a better choice than a higher TV/PBW because the risks associated with an unnecessarily low TV are lower than those associated with an unnecessarily high TV.
The methodologic and safety concerns raised by Eichacker and colleagues were addressed by the ARDS Network investigators and the Office of Human Research Protections (OHRP). The OHRP conducted a thorough investigation to assess whether the control arm was subjected to a range of TVs that conferred a disadvantage. In their published reports, the OHRP, declared that the risks to the subjects in the ARDS Network trial were minimal and reasonable in relation to the anticipated benefits given the high variability in the care of ARDS patients at the time and the lack of standard of care.17,18 In their response to Eichacker et al, the ARDS Network investigators challenged the methodology of their critic’s meta-analysis and reiterated that at the time of the study there was no standard of ventilator strategy.19 They indicated that at the time of the ARMA trial the Ppl was not used to adjust TVs in any systematic fashion and that there was disparity in physician-selected TVs and the threshold for a Ppl limit indicating equipoise in the medical community on the most appropriate approach to mechanical ventilation. A “standard control” reflective of prevailing ventilation strategy at that time did not exist and the physician’s interpretation of the preclinical data and the resulting clinical practices were highly variable. For instance, in the 5 RCTs evaluated by Eichacker et al, there were 4 different ways to calculate the TV based on PBW, IBW, dry, and measured actual body weight. The ARDS Network investigators indicated that the mean TV of patients in the traditional arm after randomization was 10 mL/kg PBW which was consistent with the prevailing clinical practice in the 1990s. They also challenged the meta-analysis conclusions of lack of efficacy due to the small sample size of the 3 nonbeneficial studies, small differences in TVs between study groups in the 2 beneficial studies, and the lack of a pooled estimate of mortality or clinical heterogeneity. In addition, the ARDS Network investigators subsequently published further data from the ARMA trial that detailed the clinical benefits of TV and Ppl reduction across the range of disease severity and Ppl on day 1 of randomization.20 They showed greater severity of disease in patients with lower respiratory system compliance and that the Ppl was an independent predictor of mortality—decreasing Ppl decreased mortality in ARDS patients, but the investigators caution that this data should not be interpreted to suggest that TV should be lowered below 6 mL/kg PBW.
A subsequent analysis of the ARMA trial examined the efficacy of a low TV ventilation strategy in patients with different clinical risk factors for ARDS.21 Even though the risk of death was significantly higher in certain subgroups of patients (sepsis, pneumonia, aspiration), a low TV strategy was equally effective across ARDS patients with different clinical risk factors, pulmonary versus nonpulmonary etiology, or infection-related versus noninfection-related conditions. The results from this study strongly support the beneficial effect of low TV ventilation in patients with diverse clinical risk factors for ARDS.
Further studies have showed a benefit of a low TV approach and a recent meta-analysis of 9 RCTs on lung-protective ventilation by Putensen et al has shed some light on the clinical benefits of this ventilatory strategy.22 Contrary to prior meta-analysis that did not focus on the comparison between lower and higher TV at similar PEEP,11,12,23 the authors analyzed data according to the effect of different lung-protective strategies: higher versus lower TV at similar PEEP, higher versus lower PEEP strategies during low TV ventilation, and lower TV and PEEP titrated greater than the LIP of the individual’s pressure-volume curve versus higher TV and lower PEEP. The authors showed that compared to higher TV ventilation at similar PEEP, lower TV ventilation-reduced hospital mortality (OR (odds risk): 0.75%, 95%CI: 0.58-0.96), a higher PEEP did not reduce hospital mortality compared with lower PEEP using low TVs but that a higher PEEP reduced the need for rescue therapy to prevent life-threatening hypoxemia. These findings support the hypothesis that the higher heterogeneity found in previous meta-analysis can be partially attributed to the inclusion of RCTs that simultaneously investigated lower TV and higher PEEP strategies. Similar to the data from Moran and Petrucci, the authors also found that a lower TV approach did not improve outcomes when higher TV ventilation results in Ppl less than 30 cm H2O. However, none of the analyses demonstrated an advantage of high TV ventilation. The authors concluded that low TV ventilation seems to be beneficial in patients with ARDS for routine clinical practice if potential side effects, such as hypercapnia and respiratory acidosis, are not contraindicated. This recent data indicate that TV is a determinant of lung-injury risk and that a lung-protective ventilator approach in ARDS limits further ventilation-induced damage locally and systemically. In the ARMA trial, low TV ventilation was associated with significant reduction in the plasma level of IL-6 and IL-8 and a more rapid attenuation of the inflammatory response by day.24 More recently, Determann et al showed a greater decrease in IL-6 levels in patients receiving low TVs as compared to a conventional approach (51 ng/mL to 11 ng/mL vs 50 ng/mL to 21 ng/mL; P = 0.01).25 The CT studies by Terragani et al identified ARDS patients in whom tidal inflation occurred largely in normally aerated compartments (more protected) and those in whom tidal inflation occurred largely in the hyperinflated compartments (less protected). They found that pulmonary cytokines were significantly lower in the more protected patients and concluded that limiting TV to 6 mL/kg PBW and Ppl to 30 cm H2O may not be sufficient in patients with larger nonaerated compartments.26 Grasso et al demonstrated that many patients with ARDS have a stress index that indicates alveolar hyperventilation while receiving PEEP according to the ARDS Network recommendations.27 These data point toward utilization of even lower TVs that may increase the likelihood of hypoxemia and other metabolic abnormalities.
Over the last few years, advances in extracorporeal technology have led to significant rise in the use of extracorporeal membrane oxygenation (ECMO) to minimize these metabolic abnormalities in ARDS.28 With the adjunctive use of ECMO, even lower TVs can be achieved while removing CO2 to correct the low-tidal-volume-induced respiratory acidosis. The recent Xtravent RCT in ARDS patients evaluated the effect of low TV (3 mL/kg PBW) using ECMO versus the ARDS Network strategy (TV of 6 mL/kg PBW) without the use of an extracorporeal device.29 Although there were no significant differences in VFDs or mortality between the study groups, a post-hoc analysis revealed significantly more VFDs at 60 days in patients with more severe hypoxemia (PaO2/FIO2 ≤ 150) who were in the ECMO group (40.9 ± 12.8 vs 28.2 ± 16.4, P = 0.033). Overall, the use of ECMO in ARDS still remains controversial with conflicting survival data and the ongoing EOLIA RCT (“ECMO to rescue lung injury in severe ARDS”) will attempt to clarify the role of ECMO in these patients.