The untapped wellspring of translational research--the auto PEEP example.

Current respiratory care practitioners (RCPs) may find it hard to believe that dynamic hyperinflation and auto-PEEP remained clinically unrecognized long after the era of positive pressure ventilation had begun. Yet, this pervasive phenomenon was largely ignored for almost 20 years, even though it occurred in virtually every patient with decompensated asthma, exacerbated chronic obstructive pulmonary disease, or high ventilation requirements. Although now viewed as an important variable to regulate, its discovery cost nothing more than the application of physiologic knowledge to an unfamiliar clinical observation. To understand why this might have been, it helps to rapidly rewind the historical record to the mindset and technology employed in the late 1970's.

In 1979 intensive care was still influenced by the "post-op" attitudes and priorities of anesthesiologists who pioneered its initial stage. Because the tacit aim of the ICU physician was to correct an acute derangement of normal physiology, the primary target for mechanical ventilation was to achieve normal blood gases, almost without regard for the pressure and ventilation costs associated with the underlying disease. The dramatic advances in understanding lung mechanics, respiratory muscle function, and ventilator-induced lung injury that occurred through the 1980's and early 1990's had not yet been brought to bear on clinical practice. Awareness of gas trapping in patients with airflow obstruction was commonplace before the initial description of auto-PEEP, but its development during mechanical ventilation and implications for cardiopulmonary management were not generally understood. Cloaking of this "occult" (hidden) PEEP was a lack of monitoring sophistication in the ventilators available. A needle gauge for pressure was the only gauge; using waveform displays and digital flow monitoring was not an option.

[ILLUSTRATION OMITTED]

In late 1979 a patient of mine hospitalized with exacerbated COPD underwent subacute respiratory decompensation that required intubation and mechanical ventilation. While immersed in my scheduled weekly experiment in the physiology lab, I took a call from my concerned resident who notified me of the dramatic fall in blood pressure that had occurred immediately after intubation, despite an unchanged chest x-ray and electrocardiogram. After a flow-directed pulmonary artery (Swan-Ganz) catheter was placed, we supported his blood pressure with fluids and vasopressors. The numbers were surprising; the measured wedge pressure exceeded 20 mm of mercury, despite a perfectly clear chest x-ray, and his cardiac output approximated a low 2.5 liters per minute. The patient's arterial blood gas indicated a nearly normal pH of 7.38, and a PaCO2 of approximately 40 mmHg, as we intended. Initially, we interpreted these data to indicate cardiogenic shock of uncertain origin. Returning to the bedside, however, my opinion quickly changed--the spirometer attached to the Puritan-Bennet MA-1 ventilator to measure expired volume indicated the appropriate tidal volume but incomplete emptying of the chest with each breath before being reset by the next inspiratory cycle. Reducing the machine's cycling frequency caused the bellows to rise to its physical limit and dump excess gas beyond its filling capacity. While I puzzled over this observation, a nurse disconnected the patient from the ventilator for airway suctioning, as was customary at the time. His monitored blood pressure rose dramatically. We then re-attached the ventilator and watched the blood pressure fall quickly again to its original level.

Primed by my experience in the physiology laboratory, I concluded that the patient was acting very much like someone might that had external PEEP applied and removed. I asked the nurse to re-measure his cardiac output and his wedge pressure while disconnected from the ventilator. To our amazement, cardiac output nearly doubled even as his wedge pressure fell from a value greater than 20 mmHg to ~ 12 mmHg--just as we might expect for release of PEEP. It was then a short logical step to link up the observations of gas trapping, continued end-expiratory filling of the bellows spirometer and the impression of unintentional end-expiratory alveolar pressure--occult or "auto" PEEP. (Auto is a prefix from the Greek word meaning "self".) Removal of the expiratory hose line that led to the bellows and occluding the expiratory port at end-expiration to stop expiratory flow caused the needle gauge of airway pressure to jump abruptly to 23 cmH2O. In retrospect, the "discovery" of auto-PEEP and the devising of a practical means for its detection took place over a span of less than 30 minutes. My fellow, Paul Pepe, and I found several patients with varying degrees of hemodynamic impairment in less than one week. Once made aware of auto-PEEP, we found it everywhere.

Mathematical models easily demonstrated the critical importance of minute ventilation in driving dynamic hyperinflation, and gave a potent rationale for pursuing a strategy of permissive hypercapnia in the ventilatory support of severe airflow obstruction--long before its use in ARDS. Indeed, reduced gas trapping may be a primary reason why a permissive hypercapnic strategy was reported in the early 1980's to dramatically reduce the mortality rate of patients ventilated for acute exacerbations of asthma. Confirmation of the importance of minute ventilation and manipulation of inspiratory flow rate was provided in the excellent work from David Tuxen and his colleagues in Melbourne, Australia, who demonstrated in a series of articles the influence of minute ventilation and inspiratory time fraction on gas trapping in that occurs in asthmatic patients.

Exploration into the therapeutic potential for intentionally generating auto-PEEP has continued for two decades. In a sense, effective use of high frequency ventilation depends on it--as may Airway Pressure Release Ventilation (APRV) when external PEEP is set to zero, as advocated by some physicians to allow efficient expiratory flow across the valve. The fundamental importance of generating auto-PEEP was also suggested by proponents of inverse ratio ventilation in the early 1980's. Inverse ratio ventilation has not been demonstrated, however, to confer any advantage over conventional ventilation with external PEEP in improving gas exchange or in avoiding hemodynamic problems.

The physiologic impact of auto-PEEP arises not from the pressure itself, but from the associated dynamic hyperinflation. Thus auto-PEEP can occur in a variety of settings not associated with air flow obstruction and is not always accompanied by adverse consequences. The active use of the expiratory muscles (as during exercise) can produce auto-PEEP with normal or even subnormal end-expiratory lung volumes. Gas trapping in dependent lung regions often occurs in the setting of ARDS, but this regional auto-PEEP is almost always relieved by modest levels of external PEEP that counter the compressive effects of the heavy lung on the dependent airways. It has also been suggested that the generation of auto-PEEP during the ARDS Network trial could have explained the relative survival advantage of patients treated with lower tidal volume but higher ventilating frequency. This point, however, remains controversial.

Clinicians continue to lean about this ubiquitous and complex phenomenon. Currently, mechanical heterogeneity is a big issue; even within the same patient, auto-PEEP varies from site-to-site and cannot always be accurately quantified from stop flow circuit occlusion. Although it has been generally recognized that extensive plugging occurs in asthmatic airways, the fact that the expiratory port occlusion method for quantifying auto-PEEP fails to detect the higher airway pressure that exists beyond these checkpoints is a relatively new observation. Plateau pressure, which is considerably easier to measure, better reflects the degree of gas trapping in such instances. Another point of investigative interest concerns the use of applied PEEP in passively ventilated patients with airflow obstruction. One of the benefits of adding external PEEP may be to counterbalance lesser values of auto-PEEP in some units while raising end expiratory alveolar pressure in others. The net effect of these actions is to help even the distribution of ventilation. Work is also ongoing to confirm the largely theoretical concern that auto-PEEP is the fundamental "feed forward link" that promotes chaotic breathing patterns in obstructed patients treated with pressure supported ventilation--independently of their voluntary efforts.

Most scientists can verify that there are few rewards to compare with the sense that bedside medical practice has been improved by a discovery stemming from one's own observations. As the auto-PEEP story exemplifies, many important observations may be missed in the rush of daily practice. Taking time to question what you do not understand sometimes will lead to implications far beyond those for the individual patient under consideration. Unfortunately for clinical practice, investigative funding has swung dramatically toward clinical trials, health services delivery, industry driven projects, and molecular biology. The true "translational" interface that can benefit patients today on a broad scale--mechanism-defining work conducted by clinician investigators using their own observations and ingenuity--has experienced a meltdown. Few physicians are doing it any more, even in academic centers. In this financially stressed environment in which we now practice, extra time for bedside research has been defined by some administrators as "unproductive", unrewarded activity and therefore discouraged.

Yet, all is not so black after all--sophisticated monitoring technology and data processing are ubiquitous--and the RCPs are continuously on the scene. Were I to advise RCPs in practice or in training, it would be to prepare well in physiology, to question what you do not understand from a base of confidence gained from that physiological knowledge, and to look very carefully--there are more phenomena like auto-PEEP waiting to be described by the brightest and most committed among you.

by John Marini MD