Why the ventilator that saves the lung can sink the right ventricle in the same breath
Airway pressure only reaches the heart after it converts to pleural pressure. That conversion is everything.
Positive pressure ventilation never touches the heart directly. Every hemodynamic effect runs through pleural pressure (Ppl), and Ppl is only a fraction of airway pressure (Paw), set by how compliant the lung happens to be. Follow that single conversion and the whole figure falls into place.
Rising Paw raises Ppl, which raises central venous pressure and collapses the pressure gradient that drives blood back to the heart.
Squeeze the venous reservoir from outside the vessel and the gradient falls even though nothing changed inside it. Pinsky's work on this remains the clearest account of why a positive pressure breath and a spontaneous breath move cardiac output in opposite directions.[1,2]
Rising Paw also raises pulmonary vascular resistance, and PVR is RV afterload. The right ventricle is a thin, compliant chamber built to eject into a low pressure circuit. It tolerates volume far better than pressure, so a PVR increase the left ventricle would shrug off can dilate and fail the RV within a handful of breaths.
The left ventricle gets the opposite deal. Rising Ppl raises the pressure surrounding the intrathoracic aorta and the LV itself, dropping LV transmural pressure and making ejection easier. This is the physiologic basis for using CPAP or PPV to treat cardiogenic pulmonary edema: the same breath that loads the right ventricle unloads a failing left one.[3]
One mechanical breath, three simultaneous effects: RV preload down, RV afterload up, LV afterload down. Minor perturbations in a healthy heart. In ARDS, where the pulmonary vasculature is already inflamed, microthrombosed, and hypoxically vasoconstricted, the RV afterload hit is the one that kills.
The hemodynamic penalty of a given airway pressure depends on where the stiffness lives, not on the ventilator.
The middle panel is the part that actually changes management. How much of Paw reaches the pleural space is governed by lung and chest wall compliance, and that determines which side of the heart takes the hit.
| Scenario | What happens | Dominant hemodynamic hit |
|---|---|---|
| Low lung compliance stiff parenchyma, "pulmonary" ARDS | Stiff lung absorbs most of Paw. Alveolar pressure climbs; Ppl barely moves. | RV afterload dominant. Venous return relatively preserved. |
| High lung compliance more compliant lung, extrapulmonary process | Pressure transmits more fully to the pleural space. | Preload dominant. RV afterload spared somewhat; venous return and LV filling suffer. |
| Elevated intra-abdominal pressure | Diaphragm pushed cephalad, chest wall compliance falls, Ppl rises independent of lung mechanics. | Preload crushed from below. High PEEP on top of high IAP is a double preload hit. |
This maps directly onto the ARDS phenotype literature. Reduced chest wall compliance (obesity, abdominal compartment syndrome, ascites, anasarca) drives Ppl up and hammers preload while relatively sparing transpulmonary pressure and RV afterload. Reduced lung compliance from consolidation or diffuse alveolar damage does the opposite: it drives transpulmonary pressure and RV afterload up while Ppl, and preload with it, stays closer to baseline.[4,5] Same PEEP, same tidal volume, opposite failure mode, depending entirely on where the stiffness lives.
Two vessel populations, opposite behavior, one nadir near functional residual capacity.
The bottom panel is the physiology that makes "just add more PEEP" a bad universal answer. Pulmonary vascular resistance is the sum of two vessel populations moving in opposite directions across the lung volume range, first characterized by Benumof and Wahrenbrock and formalized in West's zone model of the pulmonary circulation.[6,7]
Larger arteries and veins running through lung interstitium, tethered open by surrounding tissue. At low lung volumes that tethering is lost, the vessels narrow, and resistance rises. This is the atelectasis end of the curve: collapsed lung compresses its own blood supply.
Capillaries stretched across the alveolar septa. As alveoli distend toward total lung capacity, the capillary bed is squeezed flat between two overinflated air spaces, and resistance rises again. This is the overdistension end of the curve.
The nadir sits near functional residual capacity, roughly where a well recruited, non-overdistended lung lives on a properly set ventilator. PEEP that recruits collapsed lung walks the extra-alveolar side of the curve down. PEEP that overdistends already-open lung walks the intra-alveolar side back up. No PEEP setting is universally protective for the right ventricle. There is only a setting that is correct for that lung's current volume state, and overshooting it costs RV afterload from the other direction.
Acute cor pulmonale is common, ventilator-modifiable, and mortality-linked.
Vieillard-Baron's group has spent two decades characterizing what happens when this physiology goes wrong at scale. Acute cor pulmonale shows up in roughly 20 to 25 percent of moderate to severe ARDS patients even on lung protective ventilation, and it is independently associated with mortality.[8,9]
| Risk factor | Threshold flagged in the literature |
|---|---|
| Driving pressure (ΔP = Pplat − PEEP) | ≥ 18 cmH2O |
| PaCO2 | ≥ 48 mmHg |
| PaO2/FiO2 | < 150 mmHg |
Tidal volume, not plateau pressure, is the strongest driver of RV afterload in ARDS, because Vt determines how far the lung swings across the PVR curve on every breath, not just where it sits at end-inspiration.[9,10] This is the mechanistic backbone of the RV protective ventilation strategy Vieillard-Baron and Repessé proposed.[4,8]
Keep ΔP under roughly 17–18 cmH2O before chasing oxygenation with more PEEP.
PaCO2 under 60 is generally tolerated. Uncontrolled RV afterload is not.
Recruit what will actually recruit. Overshooting the PVR nadir raises afterload from the intra-alveolar side.
Any hypotension following a vent change deserves a bedside look at RV size and septal motion before assuming vasoplegia.
Recruits dorsal lung, homogenizes transpulmonary pressure, and unloads the RV as a hemodynamic intervention, not just an oxygenation maneuver.[11]
All numeric thresholds (ΔP, PaCO2, P/F cutoffs) and prevalence figures were checked against the Vieillard-Baron/Repessé/Mekontso Dessap primary series rather than taken from a secondary compilation. The PVR U-curve mechanism traces to the original Benumof/West physiology work, not a modern review's redrawing of it.