8.2 PE, Pneumonia, and Pulmonary Edema (01.N.3, 7–8)

Thromboembolism, pneumonia, and pulmonary edema are the three pulmonary processes that most often compete with the brain for blood pressure, oxygenation, and anticoagulation. Neuro ICU patients combine immobility, inflammation, indwelling lines, and — frequently — a contraindication to the very therapy PE usually requires.

Pulmonary embolism: pretest probability and diagnosis

Pulmonary embolism (PE) is venous thrombus in the pulmonary arteries. Risk in the neuro ICU is high: hemiplegia, coma, delayed chemoprophylaxis after hemorrhage, and perioperative neurosurgery. Classic symptoms (pleuritic pain, hemoptysis, isolated dyspnea) are insensitive in intubated patients. Think PE when dead-space ventilation rises, PaCO2 climbs despite a stable or increased minute ventilation, oxygenation slips without new infiltrates, or unexplained hypotension and RV strain appear.

The Wells score is a conceptual pretest tool, not a substitute for imaging in a critically ill patient who already has tachycardia and immobility. Points:

Two-level interpretation: PE unlikely ≤ 4, PE likely > 4. Traditional three-level bands are low < 2, moderate 2–6, and high > 6. In outpatients, PE-unlikely plus a negative D-dimer can stop the workup. In the ICU, D-dimer is almost always positive; go to CT pulmonary angiography (CTPA) when PE is the leading concern and the patient can travel or the scanner can come. V/Q scintigraphy is the usual alternative when iodinated contrast is unsafe and the chest radiograph is relatively clear. Bedside echo does not diagnose PE; it risk-stratifies (RV dilation, septal flattening, McConnell sign) and can justify rescue therapy in a crashing patient who cannot go to CT.

Risk strata used in guidelines:

Anticoagulation, lytics, and the limited role of the IVC filter

Start parenteral anticoagulation as soon as PE is likely and bleeding risk is acceptable — typically an unfractionated heparin infusion in the neuro ICU because it is titratable and short-acting around EVDs, craniotomy, and angiograms. Direct oral anticoagulants are maintenance drugs, not crash-ICU tools when a trip back to the operating room is plausible.

Systemic thrombolysis is for massive / high-risk PE without a prohibitive bleed risk. The labeled adult regimen is alteplase 100 mg over 2 hours (commonly 10 mg bolus then 90 mg). Cardiac-arrest dosing (for example a 50 mg bolus) appears in resuscitation algorithms but is not a reason to lyse a stable patient. CHEST guidance suggests systemic thrombolysis in hypotensive PE with acceptable bleed risk and recommends against routine lytics in PE without hypotension unless the patient later deteriorates on anticoagulation. Catheter-directed therapy or surgical embolectomy are alternatives when systemic lytics are contraindicated and local expertise exists.

Inferior vena cava (IVC) filters do not treat existing pulmonary clot. CHEST recommends a filter when anticoagulation is contraindicated, and considers one for recurrent PE despite adequate anticoagulation. PREPIC-2 showed no reduction in recurrent PE when a retrievable filter was added to anticoagulation. Filters cause thrombosis of their own; plan retrieval when the contraindication lifts.

Neuro bleed versus PE: the conflict

This is the highest-yield collision on the exam.

• Active intracranial hemorrhage, an unsecured ruptured aneurysm, and recent craniotomy or spinal surgery are standard absolute or near-absolute contraindications to systemic thrombolysis.
• Therapeutic anticoagulation after ICH, hemorrhagic transformation, or SAH is delayed and individualized. An unsecured aneurysm generally forbids full anticoagulation; once the aneurysm is secured, the question becomes EVD-tract and hematoma risk, not just the original bleed.
• Prophylactic subcutaneous heparin after ICH is a different, lower-dose decision (often after 24–48 hours of documented stability in AHA ICH discussions) and does not treat acute PE.
• When acute proximal DVT or PE occurs during a hard anticoagulation hold, an IVC filter is a bridge, not a cure. Document a date to start anticoagulation and to retrieve the filter. Catheter-based thrombus removal may be discussed for obstructive PE when lytics are off the table.
• Right-ventricular support (norepinephrine to keep CPP and coronary perfusion, cautious volume, inhaled pulmonary vasodilators if hypoxemia and RV failure dominate) can keep the patient alive while the brain and the clot are negotiated.

Worked conflict: a Hunt-Hess 4 SAH patient, aneurysm coiled yesterday, EVD in place, now hypotensive with a CTPA-proven saddle PE and a dilated RV. Systemic alteplase is ordinarily not given. Heparin may still be too early for the EVD tract. An IVC filter, hemodynamic support, and an urgent multidisciplinary decision about catheter therapy is the rational bundle — not “lyse because it is massive.”

VAP versus aspiration pneumonia

Ventilator-associated pneumonia (VAP) is pneumonia that develops after more than 48 hours of mechanical ventilation (hospital-acquired pneumonia is pneumonia after 48 hours in hospital, not necessarily ventilated). Diagnosis is clinical plus radiographic: new or progressive infiltrate plus fever, leukocytosis or leukopenia, purulent secretions, and worsening oxygenation. Quantitative cultures from bronchoalveolar lavage (BAL) or a protected specimen brush help in complex neuro ICU patients on prior antibiotics; endotracheal aspirates are easier but less specific. Empiric coverage follows local ICU ecology and risk for MRSA and Pseudomonas; de-escalate when cultures return. Prevention is a bundle: head of bed 30–45°, oral care, subglottic suction if a suitable tube is in place, SAT/SBT, and minimizing unnecessary days of ventilation.

Aspiration pneumonia follows inoculation of orogastric contents. In a supine patient the posterior upper lobes and superior lower lobes are the dependent segments; upright aspiration prefers the basal lower lobes, more often on the right. Distinguish:

A seizure or a depressed GCS at intubation can cause chemical pneumonitis that later becomes infection. Do not label every early infiltrate “VAP,” and do not withhold VAP treatment on day 4 because “it was probably aspiration at the scene.”

Neurogenic versus cardiogenic versus mixed pulmonary edema

Neurogenic pulmonary edema (NPE) is acute alveolar flooding after a CNS insult, classically aneurysmal SAH, convulsive seizure or status epilepticus, TBI, or ICH. Onset is minutes to hours (often within 2–12 hours). A catecholamine surge raises pulmonary capillary hydrostatic pressure and injures the endothelium, so the fluid can be protein-rich. Trigger zones include the hypothalamus, medulla, and insula. Many patients have a normal or near-normal left-ventricular ejection fraction; others overlap with stress (Takotsubo) cardiomyopathy, producing mixed hydrostatic and permeability edema. NPE often improves over 48–72 hours if the neurologic driver is controlled. Incidence after SAH is reported across a wide range (about 2% to over 40% depending on grade and ascertainment) and is associated with worse outcome — mostly because it marks a severe brain injury, not because the lung is independently unsurvivable.

Do not treat NPE as left-ventricular failure by default. Aggressive diuresis can drop CPP in a euvolemic SAH or TBI patient. Echo, ECG, troponin, and BNP separate the phenotypes.

Supportive care is lung-protective ventilation, treat the CNS insult (secure the aneurysm, stop seizures, control ICP), modest PEEP for hypoxemia, and euvolemia rather than empiric furosemide. If echo shows a crushed LV or RV, treat that physiology directly. If the picture persists beyond a few days with bilateral infiltrates and a P/F ≤ 300 on PEEP ≥ 5, the patient may have crossed into ARDS and should be ventilated as in Section 8.1.