Background:

Ventilator induced lung injury (VILI) remains a problem in neonatology. High frequency oscillatory ventilation (HFOV) provides effective gas exchange with minimal pressure fluctuation around a continuous distending pressure and therefore small tidal volume. Animal studies showed that recruitment and maintenance of functional residual capacity (FRC) during HFOV ("open lung concept") could reduce lung injury.

"Open lung HFOV" is achieved by delivering a moderate high mean airway pressure (MAP) using oxygenation as a guide of lung recruitment. Some neonatologists suggest combining HFOV with recurrent sigh-breaths (HFOV-sigh) delivered as modified conventional ventilator-breaths at a rate of 3/min. The clinical observation is that HFOV-sigh leads to more stable oxygenation, quicker weaning and shorter ventilation. This may be related to improved lung recruitment. This has however to our knowledge not been tested in a clinical trial using modern ventilators.

Purpose, aims:

• To compare HFOV-sigh with HFOV-only and determine if there is a difference in oxygenation expressed as a/A-ratio and/or stability of oxygenation expressed as percentage time with oxygen saturation outside the reference range.
• To provide information on feasibility and treatment effect of HFOV-sigh to assist planning larger studies. We hypothesize that oxygenation is better during HFOV-sigh.

Methods:

Infants at 24-36 weeks corrected gestational age already on HFOV are eligible. Patients will be randomly assigned to HFOV-sigh (3 breaths/min) followed by HFOV-only or vice versa for 4 alternating 1-hours periods (2-treatment, double crossover design, each patient being its own control). During HFOV-sigh set-pressure will be reduced to keep MAP constant, otherwise HFOV will remain at pretrial settings. Outcome will be calculated from normal clinical parameters including pulx-oximetry and transcutaneous monitoring of oxygen and carbon-dioxide partial pressures.

High frequency oscillatory ventilation (HFOV) has been used in neonatal respiratory care for more than three decades. HFOV provides effective gas exchange with minimal pressure fluctuation around a set mean airway pressure (MAP) functioning as a continuous distending pressure (CDP), and low tidal volume compared to conventional ventilation (CV). HFOV was therefore thought to be able to reduce the risk of bronchopulmonary dysplasia in ventilated preterm babies. However results from randomized controlled trials comparing HFOV with conventional ventilation have been conflicting and meta-analyses have not shown clear evidence that HFOV is safer or more effective than conventional ventilation neither when used as initial strategy nor as rescue strategy in preterm babies with respiratory distress syndrome (RDS). Consequently there are no absolute indications for HFOV in preterm babies and most neonatologists today use HFOV as a rescue mode when conventional ventilation is failing in the acute setting of RDS as well as in the baby with bronchopulmonary dysplasia.

Maintaining adequate functional residual capacity (FRC) together with the fraction of inspired oxygen FiO2 are the main determinants of oxygenation. The larger the FRC, the larger is the volume of available oxygen in the alveoli for gas transport. Adequate oxygen saturation (SAT) of the blood in room air or an improvement in oxygen-saturation without changing the fraction of inspired oxygen can be seen as an indirect indicator of normal or normalized FRC, and most neonatologists use oxygenation as an indirect marker for lung volume during HFOV.

The CDP or set-MAP is the main determinant of lung-aeration during HFOV. A too low MAP may cause non-homogenous aeration and atelectasis leading to atelectotrauma and redirection of airflow to more compliant alveoli leading to localized hyperinflation.

Accordingly, early animal studies showed that recruitment and maintenance of FRC during HFOV could reduce lung injury. Lung recruitment was initially achieved by superimposing conventional ventilation (CV) breaths on top of HFOV with lower MAP than used today, either as recurrent sustained inflations lasting 15-20 seconds about every 20 minute, as intermittent sigh breaths (3-5 tidal breaths pr minute) delivered as normal conventional breaths or as conventional ventilation at normal rate combined with HFOV.

Today most neonatologists perform this so-called "open lung" concept by adjusting the set-MAP using oxygenation as an indirect guide of lung recruitment. Different approaches are used explained by difficulties in direct bedside monitoring of FRC. Some initiate HFOV with MAP 2-3 cm H2O above the MAP needed during conventional ventilation subsequently adjusting MAP until the fraction of inspired O2 (FiO2) <0.25-0.6 providing no signs of over inflation of the lungs on x-ray. Others go through a more complex step-wise increase in MAP till FiO2 cannot be reduced further, and then gradually decrease MAP until FiO2 again needs to be increased to maintain a predefined SAT and then continues ventilations with a MAP set at 2 cm H2O above this point, thereby placing ventilation on the more compliant deflation limb of the pressure-volume relationship of the lung.

During HFOV, MAP may be adjusted as mentioned above. Further increase in MAP may increase FRC by increased aeration and consequently improve oxygenation. Although recent clinical trials suggest this approach is safe, it could potentially lead to generalized hyperinflation and volutrauma in addition to interfering with systemic venous return and cardiac output especially if not combined with direct monitoring of lung volume which is currently not available in routine clinical care.

Combining intermittent recruitment sigh breaths at a rate of 3-5 breaths/minute with HFOV could be an alternate way of assisting in maintaining or normalizing FRC during which MAP is only increased temporarily and intermittently. This could in theory lead to quicker weaning in MAP, less oxygen exposure and potentially reduced lung injury. A concern however could be, that the intermittent sigh breaths will lead to intermittent excessive pressures in distal airways and to excessive tidal volume and accordingly not be beneficial at all. Nevertheless the approach of combining HFOV and sigh breaths at a low rate seems to be encouraged by a number of neonatologist.

It has however to our knowledge not yet been tested in a controlled human trial.

A search on PubMed revealed no human or animal trials comparing HFOV combined with intermittent recruitment sigh-breaths at a low rate. Also no trials exploring this approach are currently registered on www.clinicaltrials.gov.

To our knowledge so far only one human trial comparing HFVO with recruitment breaths at low rate has been registered but never published (Texas Infant Star Trial).

Combining HFOV with conventional breaths has only been reported in a limited number of studies and only with focus on HFOV combined with CV at normal rate showing a possible benefit. Similar results have been reported when comparing High frequency Jet Ventilation (HFVJ) combined with CV at normal rate with HFVJ alone.

Trial rationale:

Combining intermittent recruitment sigh breaths at a low rate with HFOV could offer a further way of assisting in maintaining or normalizing FRC with only modest or no increase in MAP in alignment with the open lung concept.

A concern however could be, that the intermittent sigh breaths will lead to intermittent increased pressures in distal airways and too large tidal volume and accordingly not be beneficial at all. Despite this, the approach of combining HFOV and sigh breaths seems to be encouraged by a number of neonatologist.

It has however, to our knowledge not yet been tested in a controlled human trial. We therefore wish to conduct a controlled cross-over trial assessing the effect of HFOV combined with intermittent sigh breaths on oxygenation in ventilated neonates using oxygenation as an indirect indicator of lung recruitment.

Objective and hypothesis:

The objectives of this trial are to:

• Compare HFOV combined with intermittent recruitment sigh breaths at a rate of 3/min (HFOV-sigh) with HFOV only (HFOV-only) and examine if:

• oxygenation expressed as a/A-ratio improves with HFOV-sigh

  • a/A-ratio is a measure of oxygenation and calculated as
  • a/A-ratio = paO2/(0,95*FiO2- PaCO2),
  • paO2 and PaCO2 are measured on arterial blood if arterial access is in situ otherwise as transcutaneous values (see further down).

• stability of oxygenation improves with HFOV-sigh

  • expressed as a calculation of the percentage deviation of time spent outside the reference range for oxygen-saturation (SAT) for the given gestational age (AUC - area-under-the-curve) and comparing this with MAP and FiO2.

• Evaluate the possibility of setting up a larger randomized controlled trial We hypothesize that during HFOV-sigh the oxygenation will be improved as well at the stability of oxygenation with less time spent outside the reference range for SAT at an unchanged or lower FiO2

Trial design:

The trial is planned as a 4-period 2-treatment, double crossover clinical trial with each patient being its own control. Patients will be randomly assigned to receive HFOV-Sigh followed by HFOV-only or vice versa for four alternating 1-hours periods.

Inclusion Criteria:

• Infants at 24-36 weeks corrected gestational age

• Already ventilated with high frequency ventilation

• Requiring FiO2=21%-70% to maintain adequate oxygen saturation.

• Clinical stable

  o i.e. ventilated on current settings for more than just a few hours with stable but not necessarily normalized blood gases or transcutaneous values and oxygen requirement.

• Parent(s) or guardian able and willing to provide informed consent

Exclusion Criteria:

• Major congenital cardiovascular or respiratory abnormalities.
• The attending neonatologist responsible for the baby considers one of the ventilation modes unsuitable for the infant.
• Poor skin integrity precluding use of transcutaneous monitoring.
• Lack of parental signed written informed consent.
• Parents under 18 years of age.