End-Inspiratory Pause Enhances Ventilation Efficiency Without Affecting autoPEEP in COPD Patients During Controlled Ventilation

In patients with airflow obstruction receiving mechanical ventilation, an important objective is to reduce lung hyperinflation often using controlled hypoventilation\[1\]. At the same time, maintaining acceptable gas exchange is challenging, as major reductions in minute ventilation (VE) raises carbon dioxide (CO2) and causes respiratory acidosis, which may lead to adverse physiological consequences. Relatively prolonged end-inspiratory pause (EIP) has been shown to optimize CO2 clearance in hypoxemic mechanically ventilated patients\[2\]. Previous data suggests that, at equivalent total inspiratory-time (TI), shorter insufflations followed by EIP can enhance CO2 elimination in acute lung injury \[3\]. Adding EIP is classically discouraged in chronic obstructive pulmonary disease (COPD) because - at constant respiratory rate (RR) - prolonging inspiration reduces expiratory time (TE), and can worsen hyperinflation and impair hemodynamics\[4, 5\]. In this study, we assessed whether a breathing pattern characterized by high inspiratory flow (V ̇) plus EIP could reduce PaCO2 without inducing hyperinflation, compared with same inspiration-to-expiration time (I:E) and a ventilation pattern without EIP in patients with COPD undergoing controlled hypoventilation. Methods We performed a prospective, single-center, cross-over, randomized trial (ethical approval #10/2024) including deeply sedated and intubated adults with COPD exacerbation, PaCO2 ≥ 45 mmHg and no signs of respiratory effort. Persistent air-leaks, severe hemodynamic instability, pregnancy or intracranial hypertension were exclusion criteria. Patient's next of kin signed the informed consent. At inclusion, we collected demographic characteristics and baseline respiratory variables. A CT-emphysema score\[6\], using computed tomographies obtained within 24 hs of intubation for clinical reasons, was calculated (A.R, a pulmonologist specialized in medical imaging). Each lung was divided into 3 regions (superior, medium and inferior) based on anatomical references and were graded as no emphysema (score 0), emphysema ≤25% (score 1), ≤50% (score 2), ≤75% (score 3) and \>75% (score 4). Scores of the six regions were summed to obtain the total score, giving a minimum 0 and a maximum of 24 points. Total scores ≥ 2 are indicative of emphysema. Patients were ventilated in volume-controlled mode with square-flow waveform, tidal volume of predicted body weight (VtPBW) 6-8 ml/kg, TI 0.6-0.8 seconds, RR 10-16 breaths per minute (bpm) and I:E relationship 1:4-1:8. External positive end-expiratory pressure (PEEPext) was set to the maximum value that did not increase plateau pressure (Pplat) ≥ 1cmH2O compared to zero PEEP\[7\], and FiO2 to maintain oxygen saturation of 90-95%. Two ventilation strategies, each one applied for 30 minutes, were randomly evaluated (Figure 1A): a) ventilation without EIP, using initial ventilator settings (VentNO-PAUSE); b) ventilation with EIP (VentPAUSE), in which V ̇ was increased and 40-50% of the total inspiratory time (TI) was replaced by EIP; the remaining setting were equal to VentNO-PAUSE. At the end of each phase, we collected arterial blood gases, respiratory mechanics and basic hemodynamics. Total PEEP (i.e., PEEPtot=PEEPext + autoPEEP) and Pplat were assessed with 5-second end-expiratory, and 2-second end-inspiratory occlusions, respectively. Driving airway pressure (ΔP) was computed as Pplat - PEEPtot, normalized elastance (ERS-n) as driving pressure (ΔP)/VtPBW and inspiratory airway resistance (Raw) as (Peak pressure \[Ppeak\] - Pplat) / V ̇. We hypothesized that ventilation efficiency would be better during VentPAUSE, and wanted to evaluate whether this strategy would have allowed to reduce VE while keeping the same CO2 obtained without EIP. Accordingly, and assuming a constant CO2 production, we calculated the predicted VE during VentPAUSE to maintain the same CO2 measured during VentNO-PAUSE with the formula : Predicted VE\_( (Vent\_PAUSE))=█(〖PaCO\_(2 )〗\_((Vent\_PAUSE ) )@ )/〖PaCO\_2〗\_((Vent\_(NO-PAUSE) ) ) × VE used during the study phases \[8\] This allowed to calculate to what extent it would have been possible to reduce VE (VEpred) and RR (RRpred) using the experimental strategy. We additionally calculated the predicted prolongation of expiration using the new RRpred as: TE-pred = (60/RRpred) - TI. We calculated the ventilatory ratio as an indice reflecting physiological dead-space ventilation (VD/VT)\[9\].