19.3 Malignant Hemispheric Infarction and Basilar Artery Occlusion (03.C.1–2)

19.3 Malignant Hemispheric Infarction and Basilar Artery Occlusion

Quick Answer: Malignant MCA edema peaks around days 2–5. In patients ≤60 years, decompressive hemicraniectomy within 48 hours (pooled DECIMAL / DESTINY / HAMLET) cuts death by about half and doubles the chance of mRS 0–3. In patients ≥61 years, DESTINY II saves lives but no one in that trial reached mRS ≤2 — say that out loud. Basilar occlusion: ATTENTION (to 12 h) and BAOCHE (6–24 h, PC-ASPECTS ≥6) roughly doubled mRS 0–3 (~46% versus 23–24%). Coma and a late clock do not auto-exclude if the brainstem is not already infarcted.

A complete MCA infarct can look “stable” at hour 6 and herniate at hour 40. A basilar occlusion can look like “coma, too late” at hour 14 with a still-dark pons. Independent OpenExamPrep teaching here covers malignant hemispheric infarction and basilar-artery occlusion as Neurocritical care diseases in the ABPN Content Specifications. This guide is not an ABIM or ABPN product.

What “malignant” means

Malignant hemispheric infarction is space-occupying middle cerebral artery (MCA) (often plus ACA) infarction with life-threatening edema, midline shift, and herniation. Medical mortality without decompression historically approached 70–80%. Osmotherapy, head-of-bed, brief hyperventilation as a bridge, and ICU care treat ICP; they do not replace bone removal when the hemisphere is a tight mass. Peak swelling is usually 48–96 hours, sometimes later in thrombolyzed or recanalized patients who still have a large core. The surgical trials enrolled patients with declining consciousness (NIHSS item 1a ≥1), not only those already with a unilaterally fixed pupil.

Radiographic flags used in trials and guidelines: infarction of >50% of the MCA territory on CT, or DWI volume >145 cm³ (DECIMAL). Younger brains have less atrophy and less room. A 35-year-old with a complete MCA is a decompression candidate far more often than an 85-year-old with the same hypodensity and a spacious extra-axial space — age is not only a disability statistic; it is intracranial compliance.

DECIMAL, DESTINY, HAMLET: ≤60 years, <48 hours

Three European randomized trials tested early hemicraniectomy versus medical ICU care:

• DECIMAL (France): DWI-defined large MCA infarcts.
• DESTINY (Germany).
• HAMLET (Netherlands): allowed surgery up to 96 hours; the <48 hour subgroup is the one that matches the pooled benefit.

The pooled analysis (Vahedi and colleagues, Lancet Neurology 2007) of patients ≤60 years operated within 48 hours showed:

• Survival about 78% versus 29% (absolute mortality reduction on the order of 50%).
• mRS 0–3 (walking with some independence in daily life) about 43% versus 21% of all randomized patients.
• Many additional survivors live at mRS 4 (unable to walk unassisted). Surgery does not manufacture a normal hemisphere. Families who hear “we can save your spouse’s life” must also hear “they may need 24-hour care.”

Technique that matches the trials: a large hemicraniectomy (diameter ≥12 cm), temporal decompression so the uncus has somewhere to go, duraplasty / open dura — tight primary dural closure defeats the operation. Leave the bone flap out until swelling subsides (cranioplasty later). Continue medical ICP care; hypothermia added to hemicraniectomy has not earned a routine recommendation.

Timing: the benefit is early surgery, before the brainstem is irreversibly compressed. HAMLET’s delayed-surgery overall result is the cautionary tale: waiting for “the scan to look really bad” can wait past the window. IV alteplase earlier in the course is not a contraindication to later decompression (ENCHANTED post-hoc and surgical series).

DESTINY II and the older adult: shared decisions

DESTINY II (Jüttler and colleagues, NEJM 2014) randomized patients 61 years or older. Hemicraniectomy improved the proportion with mRS 0–4 versus 5–6, almost entirely by converting deaths into severe disability. Six-month mortality was about 33% with surgery versus 70% with medical care in the trial report. The 2026 AHA/ASA surgical synopsis cites 12-month mortality around 42% surgical versus 76% medical, mRS 3 in 6% of the surgical group, and no survivor with mRS ≤2 in either arm. That is the sentence to put on the table: we can often keep this person alive; we cannot promise they will walk or live independently. Some families value survival in any form; some do not. Either choice can be right. The wrong choice is performing a 15 cm craniectomy at 03:00 without a decision-maker who has heard DESTINY II.

The 2026 AHA/ASA text still treats early decompression ≤60 years as the strongly supported functional-and-survival intervention, and treats >60 years as a survival intervention without a demonstrated gain in favorable recovery. Do not hide that distinction behind “the guidelines say operate.”

Dominant versus nondominant hemisphere

Trial populations included left (usually language-dominant) and right hemispheres. Aphasia is not an official exclusion. Right-hemisphere malignant infarcts add neglect, anosognosia, and agitation that families sometimes experience as a different person. Left-hemisphere infarcts add global aphasia that families experience as lost communication. Neither is a “quality of life so low that surgery is unethical by default.” Both require a specific conversation: What would this person have said about living without speech? Without walking? With a feeding tube? Document the prestroke mRS, occupation, and prior directives. Operate on the hemisphere that meets DECIMAL/DESTINY anatomic and age criteria when the decision-maker, informed, says yes. Do not reserve surgery only for the nondominant side because you personally fear aphasia more than neglect.

After decompression, expect a large CSF/skin flap bulge, seizures, wound issues, and a later cranioplasty infection risk. Hydrocephalus can appear. Rehabilitation starts when ICP physiology allows, not when the flap looks cosmetic.

Basilar artery occlusion: do not abandon the clock alone

Basilar artery occlusion (BAO) presents as coma, locked-in syndrome, internuclear ophthalmoplegia, pinpoint or mid-position pupils, quadriparesis, or “unexplained unresponsiveness” after a prodrome of vertigo and diplopia. Mortality and locked-in survival without reperfusion are high. IVT still belongs in the 0–4.5 hour eligible patient; it is not a substitute for EVT when the basilar is occluded.

Early randomized signals were messy. BEST (China) and BASICS (international, many milder NIHSS, high IVT rates, 0–6 hour emphasis) did not show a statistically significant primary-endpoint win, with crossovers and enrollment problems. Then ATTENTION and BAOCHE (both 2022, NEJM) enrolled sicker patients with better imaging discipline:

• ATTENTION (Tao and colleagues): BAO within 12 hours. EVT versus medical care: 90-day mRS 0–3 46% versus 23%. sICH 5% versus 0% (not statistically decisive). Median NIHSS was high — this was not a dizziness-only cohort.
• BAOCHE (Jovin, Li, and colleagues): randomization 6–24 hours after onset, posterior-circulation ASPECTS (PC-ASPECTS) ≥6, and exclusion of extensive brainstem infarct (pons-midbrain index ≥3 after protocol tuning). mRS 0–3 46% versus 24%; stopped early for efficacy. Few patients had NIHSS 6–9; after an amendment the floor was NIHSS ≥6, but the evidence mass is NIHSS ≥10.

The 2026 AHA/ASA posterior-circulation recommendation is Class 1 for EVT within 24 hours in BAO with prestroke mRS 0–1, NIHSS ≥10, and PC-ASPECTS ≥6. That is the exam sentence. Milder NIHSS (<10) and devastated PC-ASPECTS (<6) remain evidence-thin. PC-ASPECTS is a 10-point posterior score (cerebellar hemispheres, thalami, PCA territories, midbrain, pons) — analogous in spirit to anterior ASPECTS, not interchangeable with it.

Do not refuse EVT only because last known well is 18 hours if the pons is still dark, PC-ASPECTS is 8, and CTA shows a mid-basilar cutoff. That is BAOCHE country. Do refuse magical thinking when MRI already shows a completed bilateral pontine infarct, no collaterals, and a family hoping the catheter will reverse brain death. Coma at presentation is why you call the interventionalist, not why you consign the patient to a dark room. Distinguish locked-in (ventral pons, aware, vertical gaze/blinks) from tegmental coma; both can be basilar, and only one can hear your prognosis speech.

Cerebellar infarcts that swell are a different operation (suboccipital decompression ± EVD for fourth-ventricle obstruction) taught with posterior fossa mass effect — do not perform a supratentorial hemicraniectomy for a tight cerebellum.

Worked stems

A 52-year-old with complete left MCA infarct, NIHSS 18, drowsy at 30 hours, NCCT >50% MCA, right arm still withdraws: offer hemicraniectomy now. Do not wait for a blown pupil.

A 74-year-old with the same scan, previously independent: DESTINY II conversation. If they would not accept mRS 4–5 survival, medical ICU care and palliation are legitimate. If survival is the goal, operate early, not after herniation.

A 60-year-old found comatose, last known well 16 hours, NIHSS 28, CTA mid-basilar occlusion, PC-ASPECTS 8: EVT (BAOCHE/ATTENTION physiology, 2026 Class 1 envelope). Do not stop because the patient is comatose or because 16 > 6.

A 60-year-old with the same clock, MRI pons already white throughout, PC-ASPECTS 2: imaging-unfavorable; EVT is unlikely to match trial benefit. Say so.

Exam traps

Waiting 72 hours “to be sure they need surgery” in a 50-year-old. Promising an older DESTINY II patient that they will walk independently. Refusing dominant-hemisphere decompression by policy. Declaring all basilar occlusions futile at 8 hours. Calling locked-in patients comatose and stopping. Skipping IVT in a 2-hour basilar patient because “they need the lab.” Independent practice items at /practice/abim-neurocritical-care drill these conversations.