PHARLAP

Maximal Recruitment Open Lung Ventilation in Acute Respiratory Distress Syndrome (PHARLAP): A Phase II, Multicenter, Randomized, Controlled Trial

Hodgson CL et al. for the PHARLAP Study investigators. AJRCCM July 2019; DOI: 10.1164/rccm.201901-0109OC

Clinical Question

  • Does an open lung ventilation strategy which includes maximal recruitment maneuvers (RMs) and positive end-expiratory pressure (PEEP) titration improve outcomes compared to traditional lung-protective ventilation in patients with moderate – severe acute respiratory distress syndrome (ARDS)?

Background

  • In ARDS, the injured lung is often conceptualized as containing two physiologically distinct regions: a small aerated “baby lung” and a consolidated gasless compartment (DOI 10.1007/s00134-015-4200-8)
  • Interventions such as RMs, high PEEP, and prone positioning may increase the size of the baby lung through alveolar recruitment 
  • “Open lung ventilation” describes a general approach to mechanical ventilation that aims to “open the lung and keep it open.” RMs are used to maximize alveolar recruitment. PEEP is then titrated to prevent downstream derecruitment. This strategy is typically paired with low-tidal-volume ventilation to limit volutrauma
  • The benefits of such an approach, and the best means to achieve it, remain a source of debate (DOI https://doi.org/10.1007/s00134-019-05734-7)
  • The use of RMs in patients with moderate – severe ARDS received a conditional recommendation in recent consensus guidelines on mechanical ventilation ARDS (https://www.atsjournals.org/doi/full/10.1164/rccm.201703-0548ST)

Design

  • Multicenter prospective randomized phase II clinical trial
  • Blinding not performed given the nature of the intervention
  • Randomization performed using a web-based system. Patients randomized in a 1:1 fashion with permuted blocks, stratified for site and for cause of ARDS
  • Power calculation
    • Investigators estimated 340 patients would be required to provide 80% power to detect a difference equal to 33% of a standard deviation (3 ventilator free days [VFD]) with a 2-sided P value of 0.05
    • This power calculation allowed for up to a 5% withdrawal or loss to long-term follow up
  • Analysis was performed using an intention-to-treat approach

Setting

  • 35 intensive care units (ICUs) in 5 countries (Australia, Ireland, Kingdom of Saudi Arabia, New Zealand, and the United Kingdom)
  • October 2012 to September 2017

Population

  • Inclusion:
    • Intubated and receiving mechanical ventilation
    • Duration of mechanical ventilation ≤ 72 hours for a diagnosis of moderate – severe ARDS as defined by the Berlin conference criteria
  • Exclusion:
    • < 16 years of age
    • > 72 hours since the diagnosis of ARDS
    • Mechanical ventilation > 10 days
    • Barotrauma
    • Chest trauma
    • Active bronchospasm or history of significant obstructive lung disease
    • Clinical suspicion of significant restrictive lung disease
    • Concern for elevated intracranial pressure
    • Unstable cardiovascular status
    • Pregnant
    • Receiving extracorporeal membrane oxygenation (ECMO)
    • Use of high frequency oscillatory ventilation (HFOV)
    • Imminent death
    • Refusal of the treating physician
  • Trial enrollment
    • 835 patients assessed for eligibility
    • 175 eligible but not enrolled
      • The most common reason not to be enrolled was clinician preference (50%)
    • 115 randomized (58 in intervention arm and 57 in the control arm)
    • 1 patient withdrew consent in each arm
    • Primary analysis includes 57 patients in the intervention arm and 56 patients in the control arm
    • No patients were lost to follow-up
  • Baseline characteristics were well matched
    • Age: ≈54 years
    • Cause of ARDS: pneumonia (61%), extra-pulmonary (21%)
    • Time since onset of ARDS: ≈18 hrs
    • Vasoactive medicines: 82%
    • CRRT: 19%
    • Neuromuscular blockade: 57%
    • Prone positioning: 4%
    • PaO2/FiO2: ≈129
    • Tidal volume (VT): ≈7 mL/kg PBW
    • Pplt: ≈28 cmH2O
    • PEEP: ≈13 cmH2O
    • Respiratory system compliance: ≈ 31 mL/cmH2O

Intervention (PHARLAP strategy)

  • Following randomization, patients placed on pressure controlled ventilation, Pinsp 15 +/- 3 cmH2O to achieve a goal V4 – 6 mL/Kg PBW. PEEP kept at pre-randomization level
  • In the absence of contraindications, a combined open lung procedure (COLP) was performed at least daily until day 5 (or until day 3 if there was no improvement in static lung compliance from baseline – day 3)
  • The COLP consisted of the following
    • Staircase recruitment maneuver (SRM)
      • FiOadjusted to maintain arterial saturation 90 – 92% for 15 min
      • PEEP increased to 20 cmH2O
      • Step-wise increases in PEEP every 2 minutes to 30 cmH2O and then 40 cmH2O
      • Procedure terminated if evidence of hemodynamic instability, SpO2< 85%, or new air leak through existing chest tube. If the clinician felt the patient was under-resuscitated, intravenous fluids and/or vasopressors could be given and the procedure repeated
      • The PEEP level one step below the PEEP level at which SRM was abandoned was considered to be the maximum tolerated PEEP for that day
    • PEEP titration
      • After completing SRM, PEEP was then reduced to 25 cmH2O (17.5 cmH2O if SRM was abandoned at 20 cmH2O)
      • At 3 minute intervals, PEEP was decreased in increments of 2.5 cmH2O to a minimum of 15 cmH2O
      • De-recruitment PEEP was defined as the PEEP level at which the SpOfirst decreased by ≥ 2%
      • PEEP not reduced below 15 cmH2O if de-recruitment did not occur
    • Brief recruitment maneuver (BRM)
      • 2 minute RM with PEEP set at maximum tolerated PEEP achieved during SRM
      • PEEP then returned to PEEP level 2.5 cmH2O above de-recruitment PEEP (this PEEP level was considered the daily optimal PEEP)
    • Subsequent management
      • Pinsp titrated to achieve Pplt 25 – 28 cmH2O and a V4 – 6 mL/Kg PBW
      • FiO2 titrated to SpO90 – 95%
      • After 1 hr, a decrease in SpOwas considered a strong indication to repeat a BRM
      • Rescue therapies for hypoxemia (prone positioning, inhaled nitric oxide, HFOV, ECMO, etc) could be considered for PaO2< 60 mmHg or a SpO< 90% on FiO≥ 0.8 if no improvement with or contraindication to BRM
    • Contraindications to COLP
      • Mean systolic BP ≤ 60 mmHg despite fluids and vasopressors
      • Active air leak through chest tube
      • Radiographic evidence of pneumatoceles, subpleural cysts, or pericardial/mediastinal emphysema
      • Subcutaneous emphysema not related to trauma or procedures
      • Supraventricular tachycardia with a mean systemic blood pressure ≤ 70 mmHg or any ventricular tachycardia

Control

  • Patients ventilated in volume control mode
  • V4 – 6 mL/Kg PBW to achieve Pplt < 30 cmH2O
  • PEEP/FiOset using the ARDSnet low PEEP table (http://ardsnet.org/files/ventilator_protocol_2008-07.pdf)
  • RR ≤ 35 breaths/min
  • pH 7.3 – 7.45
  • Rescue therapies for hypoxemia could be considered for PaO2< 60 mmHg or a SpO< 90% on FiO≥ 0.8 and a PEEP ≥14 cmH2O
  • RMs not permitted

Weaning Protocol Common to Both Groups

  • Patients considered “ready” to wean when:
    • SpO≥ 90% on FiO≤ 0.4 for ≥ 6 hrs
  • Both T-piece and PS trials allowed at clinician discretion
  • Patient assessed clinically for possible liberation after a 30 minute SBT

Outcome

  • Primary outcome: Ventilator Free Days at day 28
    • Intervention 16 (IQR 0 – 21) vs control 14.5 (0 – 21.5)
    • P value = 0.95
  • Impact of the intervention
    • Protocol adherence
      • 97% of patients in intervention arm received at least 1 COLP
      • 21% of patients in the control arm received a non-protocolized RM
    • COLPs in intervention arm
      • 102 performed
        • 30% associated with a clinically significant event
          • Transient desaturation: 28%
          • Hypotension requiring increased vasopressors: 13%
        • 61% completed SRM to 40 cmH2O
        • 90% performed by primary ICU consultant
        • 2 SRMs abandoned and not repeated due to both desaturation and hypotension
    • Number of RMs in intervention arm between days 1 – 5
      • 1 = 50%; 2 = 22%; 3 = 19%; 4 = 2%; 5 = 3%
    • Respiratory variables
      • Day 1: Intervention arm with
        • Higher PEEP: 16 cmH2O vs 11 cmH2O, P = <0.0001
        • Higher Pplt: 28 cmH2O vs 25 cmH2O, P = 0.005
        • Lower driving pressure: 11 cmH2O vs 13 cmH2O, P = 0.038
        • Higher PaO2/FiO2: 203 vs 132, P < 0.0001
        • Higher PaCO2: 57 mmHg vs 50 mmHg, P = 0.002
        • Lower pH: 7.28 vs 7.35 P = 0.0001
      • Day 3: Intervention arm with
        • Higher PEEP: 13 cmH2O vs 11 cmH2O, P = 0.01
        • Higher Pplt: 27 cmH2O vs 25 cmH2O, P = 0.04
        • Higher PaO2/FiO2: 211 vs 150, P < 0.0001
      • Day 7: No significant difference in any respiratory variable between groups
  •  Secondary Outcomes
    • There was no statistically significant difference between groups in any secondary outcome including:
      • death in the ICU
      • death in the hospital
      • 28-day mortality, 90-day mortality, 6-month mortality
      • duration of ventilation
      • ICU length of stay
      • hospital length of stay
  • Safety and exploratory outcomes (intervention vs control)
    • New cardiac arrhythmia: 30% vs 12.5% (P = 0.03)
    • Bradycardia: 6.9% vs 5.4% (P = 0.99)
    • Severe hypotension: 34.5% vs 21.4% (P = 0.12)
    • New inhaled nitric oxide: 10.3% vs 28.6% (P = 0.03)
    • New prone positioning: 6.9% vs 12.5% (P = 0.02)
    • New ECMO: 1.7% vs 12.5% (P = 0.03)
    • New NMB: 27.6% vs 25% (P = 0.47)
    • Refractory acidosis: 15.5% vs 7.1% (P = 0.16)
    • Barotrauma: 5.2% vs 10.7% (P = 0.32)
    • Pneumothorax requiring chest tube: 5.2% vs 5.4% (P = 0.99)
    • Pneumothorax requiring drainage ≤ 7 days: 5.2% vs 3.6% (P = 0.99)
    • Barotrauma ≤ 7 days: 5.2% vs 5.4% (P = 0.99)
  • Pre-specified subgroup analysis
    • No significant difference between groups when considering moderate vs severe ARDS, diffuse vs focal ARDS, or serum biomarkers (IL-6 and IL-8)
  • Pre-specified exploratory analysis
    • Looking only at patients in the intervention arm, patients who had increased lung compliance from day 1 – 3 had increased VFDs compared to non-responders
  • Post-hoc meta-analysis
    • Authors identified 4 RCTs of maximal open lung approach that used SRMs and PEEP titration
    • No difference in hospital mortality
      • N=1,341
      • OR 1.11, 95% CI (0.89 – 1.39, P = 0.35)
    • Increased risk of barotrauma
      • N=1,344
      • OR 1.74, 95% CI (1.01 – 2.98, P = 0.04)

Authors’ Conclusions

  • In patients with moderate – severe ARDS, an open lung strategy that included maximal RMs and PEEP titration compared to conventional lung-protective ventilation did not improve VFDs or mortality. This intervention was associated with increased cardiovascular events but reduced use of rescue therapies for hypoxemia

Strengths

  • Multinational group of study centers increases generalizability of the results
  • Intention-to-treat analysis
  • Baseline characteristics were well balanced between the two study arms
  • High adherence to trial protocol in the intervention arm
  • No patients lost to follow-up

Weaknesses

  • The trial was stopped short of its enrollment goal as the study management committee felt equipoise to continue the trial was lost after the publication of the ART trial (https://www.thebottomline.org.uk/summaries/icm/art/) which showed a signal of harm in patients randomized to receive aggressive RMs paired with PEEP titrated to best respiratory system compliance
  • 21% of patients in the control arm received a non-protocolized RM
  • 90% of COLPs were performed by ICU consultants. This level of involvement by senior members of the ICU team likely does not reflect usual practice at many centers
  • Mechanical ventilation in the control arm was guided by the ARDSnet low PEEP/high FiO2 table. It is difficult to know if the improved respiratory variables in the intervention arm (e.g. higher PaO2/FiO2, lower driving pressure, less need for rescue therapies) were simply a consequence of the patients in the intervention arm receiving higher PEEP and unrelated to the COLP. A similar concern was raised with the findings of the first EPVent study (https://www.thebottomline.org.uk/summaries/icm/epvent2/)
  • The low utilization of prone positioning is at odds with the strong recommendation for proning in patients with severe ARDS (https://www.atsjournals.org/doi/full/10.1164/rccm.201703-0548ST)

The Bottom Line

  • This study, when considered with the ART trial and the updated post-hoc meta-analysis performed by the authors, suggests that an open lung ventilation approach does not improve patient-centered outcomes when used broadly in patients with moderate – severe ARDS
  • While the striking signal of harm seen in the intervention arm of ART was not seen in this study (perhaps due to protocol differences or the high level of ICU consultant involvement during COLPs in PHARLAP), the results do suggest that aggressive RMs are associated with adverse hemodynamic effects
  • I agree with the authors’ hypothesis that there are likely to be sub-phenotypes of ARDS patients who may uniquely benefit from an open lung approach. However, before searching for a small subpopulation who may benefit from an intervention of unclear efficacy, I think the Critical Care community as a whole should focus on efforts to increase utilization of evidence-based strategies like prone positioning (https://www.thebottomline.org.uk/summaries/icm/proseva-study-group-prone-positioning-in-severe-acute-respiratory-distress-syndrome/)
  • I will continue to not use an open lung ventilation strategy as a routine part of my management of patients with moderate – severe ARDS

External Links

Metadata

Summary author:James M. Walter (@macwalterMD)
Summary date: 3rd September 2019
Peer-review editor: Steve Mathieu

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.