8.1 Pulmonary Physiology and ARDS (01.N.1–2)

Pulmonary failure is a daily problem in the neuro ICU. Arterial oxygen and carbon dioxide set cerebral blood flow, intracranial pressure (ICP), and secondary ischemic risk, so a ventilator change is also a brain change. This section covers the gas-exchange physiology that makes those trade-offs predictable, then the ARDS bundle that still drives most exam items: the Berlin definition, low tidal volume, plateau-pressure limits, PEEP/FiO2 tables used as titration recipes, prone positioning from PROSEVA, and the ROSE nuance that early neuromuscular blockade is not automatic.

West zones of the lung

John B. West described regional pulmonary blood flow as the interaction of alveolar pressure (Palv), pulmonary arterial pressure (Part), and pulmonary venous pressure (Pven). In an upright person the gradient runs from apex to base. In a supine ICU patient the same physics runs from nondependent (ventral) to dependent (dorsal) lung — which is why prone positioning recruits dorsal units rather than “the bases.”

Raising PEEP or dropping cardiac output can convert Zone 3 into Zone 1: more dead space, a higher PaCO2 for the same minute ventilation, and a lower mixed-venous PO2 if stroke volume falls. An “oxygenation win” from PEEP can still harm the brain if PaCO2 rises or mean arterial pressure falls and cerebral perfusion pressure (CPP) follows it down.

Shunt versus dead space

Ventilation–perfusion (V/Q) mismatch is a spectrum, not two diseases. Shunt is perfusion without ventilation (V/Q → 0). Dead space is ventilation without perfusion (V/Q → ∞). Low-but-not-zero V/Q units behave like shunt for oxygen but still exchange some CO2; high-but-not-infinite V/Q units waste ventilation.

The shunt fraction is Qs/Qt = (CcO2 − CaO2) / (CcO2 − CvO2). Dead-space fraction by the Bohr equation is Vd/Vt = (PaCO2 − PECO2) / PaCO2; a normal value is about 0.30 and it climbs in PE and in overdistended ARDS. Exam items usually ask you to match the pattern (hypoxemia that ignores oxygen versus hypercapnia after a PE or a PEEP jump), not to grind the algebra.

Berlin definition of ARDS

The 2012 Berlin definition remains the diagnostic frame most vignettes still use. ARDS is acute hypoxemic respiratory failure with four required pieces:

1. Timing — within 1 week of a known clinical insult or of new or worsening respiratory symptoms.
2. Chest imaging — bilateral opacities not fully explained by effusions, lobar collapse, or nodules.
3. Origin of edema — respiratory failure not fully explained by cardiac failure or fluid overload. Use echocardiography when there is no clear risk factor.
4. Oxygenation — PaO2/FiO2 (P/F) measured on at least 5 cm H2O of PEEP (CPAP is allowed for mild disease).

A 2024 global definition expanded ARDS to include patients on high-flow nasal oxygen, allowed SpO2/FiO2 criteria, and described resource-limited settings without a PEEP requirement. Board-style stems still typically hand you a P/F ratio on PEEP ≥ 5 and expect Berlin severity. Common neuro ICU precipitants are aspiration at intubation, pneumonia, transfusion, sepsis, and neurogenic or mixed pulmonary edema that then meets Berlin criteria if it persists.

Low tidal volume and plateau pressure

The ARDS Network ARMA trial (NEJM 2000) randomized 861 adults with early ALI/ARDS to a starting tidal volume of 6 mL/kg predicted body weight (PBW) with a plateau-pressure ceiling of 30 cm H2O, versus 12 mL/kg PBW with a 50 cm H2O ceiling. Death before discharge home breathing unassisted was 31.0% versus 39.8% (P = 0.007), and ventilator-free days improved. That is the strongest mortality evidence in ARDS ventilation.

Tidal volume is not based on actual body weight. ARDSNet PBW:

• Men: 50 + 0.91 × (height in cm − 152.4)
• Women: 45.5 + 0.91 × (height in cm − 152.4)

Worked example: a 170 cm woman has PBW ≈ 45.5 + 0.91 × 17.6 ≈ 61.5 kg. Target tidal volume at 6 mL/kg is about 370 mL. If she weighs 90 kg, 6 mL/kg actual weight is 540 mL — enough to overdistend the small remaining “baby lung.” Measure plateau pressure with an end-inspiratory pause on a passive patient. If plateau exceeds 30 cm H2O, reduce tidal volume toward 4 mL/kg PBW. Driving pressure (plateau − PEEP) near or below 15 cm H2O tracks mortality in observational work (Amato, NEJM 2015); it supplements, rather than replaces, the 6 mL/kg starting rule.

The FACTT trial showed that a conservative fluid strategy increased ventilator-free days in ARDS once shock was resolving — relevant when osmotherapy, SAH euvolemia, and lung water compete in the same patient.

PEEP and FiO2 tables, conceptually

ARDSNet published lower-PEEP / higher-FiO2 and higher-PEEP / lower-FiO2 tables. They are matched titration recipes, not proofs that one PEEP is “correct.” You move PEEP and FiO2 together to keep oxygenation in a protocol band (classically PaO2 55–80 mm Hg or SpO2 88–95%) while watching plateau pressure, dead space, blood pressure, and RV function.

Pick a table and stay on it; do not mix rows from both. High PEEP can flatten the right ventricle, raise Vd/Vt, and drop CPP if MAP falls. The ART trial of aggressive stepwise recruitment plus high PEEP increased mortality — recruitment maneuvers are not routine rescue.

Prone positioning: PROSEVA

PROSEVA (NEJM 2013) enrolled intubated patients with P/F ≤ 150 mm Hg, FiO2 ≥ 0.60, and PEEP ≥ 5, and used prone sessions of at least 16 consecutive hours. Twenty-eight-day mortality was 16.0% prone versus 32.8% supine in experienced European centers that also used low tidal volume (and, in most patients, neuromuscular blockade during the turn). Proning recruits dorsal lung, improves V/Q matching, and can unload the right ventricle.

Neuro ICU constraints are the exam twist:

• Uncontrolled ICP, an unstable cervical or thoracolumbar spine, an unprotected craniectomy bone flap, a tenuous airway, and EVD, bolt, or lumbar-drain hardware that cannot be secured are relative or absolute barriers.
• The turn itself can spike ICP via abdominal pressure, neck rotation that impairs jugular drainage, and coughing.
• If you prone a monitored patient, keep the neck neutral, raise the head as much as the position allows, lock the EVD to a stated reference, and watch ICP and CPP through the turn — not only after it.

Neuromuscular blockade: ACURASYS versus ROSE

ACURASYS (2010) reported a 90-day mortality benefit from 48 hours of cisatracurium in early severe ARDS (P/F ≤ 150), with both groups deeply sedated. ROSE (PETAL Network, NEJM 2019) compared early continuous cisatracurium plus deep sedation with a lighter-sedation usual-care strategy in moderate-to-severe ARDS. Ninety-day in-hospital mortality was 42.5% versus 42.8% (stopped early for futility), and the cisatracurium group had more cardiovascular adverse events. Contemporary teaching: early neuromuscular blockade is not automatic. Use it when severe asynchrony, double-triggering, or prone positioning cannot be managed with sedation and ventilator adjustment. Paralysis erases the neurologic examination and can mask seizures — a high price in the neuro ICU.

Hypercapnia versus the injured brain

Lung-protective ventilation often raises PaCO2. Carbon dioxide is a potent cerebral vasodilator; hypercapnia increases cerebral blood flow and can raise ICP. Permissive hypercapnia that is acceptable in a medical ICU ARDS patient may be unacceptable after TBI, ICH, or a tight posterior fossa. Do not treat herniation with prolonged hypocapnia: the Brain Trauma Foundation 4th edition reserves brief hyperventilation as a bridge and warns against PaCO2 < 25 mm Hg. Keep tidal volume at 6 mL/kg PBW and look for other ways to control CO2: modest respiratory-rate increases if auto-PEEP allows, reduced circuit dead space, treating shock, and, in refractory hypoxemic/hypercapnic ARDS, extracorporeal support rather than 12 mL/kg volumes. Oxygen targets still matter: hypoxia also increases cerebral blood flow and can worsen ICP, so the goal is lung protection without abandoning the brain.