20.2 Cerebellar Hematoma (03.D.2)
Key Takeaways
- Cerebellar hematoma larger than 3 cm in diameter (or about 15 mL), brainstem compression, or obstructive hydrocephalus is a surgical emergency; suboccipital evacuation is lifesaving.
- Do not wait for coma: posterior-fossa deterioration can be abrupt, and patients who reach coma before decompression fare worse.
- An external ventricular drain alone does not treat direct brainstem compression and can precipitate upward herniation if the fourth ventricle is obstructed.
- Cerebellar surgery evidence is distinct from STICH/STICH II hemispheric data; those trials excluded posterior-fossa clots and must not be used to justify watchful waiting.
A cerebellar hematoma is not “just ICH in a different ZIP code.” The posterior fossa is a small box: the clots sit against the brainstem, and even modest expansion can flatten the fourth ventricle, obstruct cerebrospinal fluid, and produce obstructive hydrocephalus or tonsillar herniation. Patients can walk into the emergency department with vertigo and then become apneic. Blueprint item 03.D.2 tests whether you will decompress before coma, and whether you understand that external ventricular drainage (EVD) alone is not enough when the problem is direct brainstem compression.
Why the location changes the operation
Supratentorial deep clots destroy internal capsule and thalamus; removing them often does not restore those pathways, which is why STICH and STICH II were able to randomize surgeons who were genuinely uncertain. Cerebellar clots, by contrast, frequently kill by mechanical compression of a still-viable brainstem and by fourth-ventricle obstruction. If you remove the mass, the brainstem can resume function. That is why neurosurgical series and the 2022 AHA/ASA spontaneous ICH guideline treat deteriorating cerebellar hemorrhage as an operation that reduces mortality, even though there is no STICH-style randomized trial in this location—those trials excluded posterior-fossa hemorrhages.
Typical hypertensive cerebellar ICH arises in the region of the dentate nucleus. Lobar cerebellar or hemispheric-plus-vermian clots from CAA, anticoagulation, or arteriovenous malformation follow the same surgical rules once mass effect is present; the vascular workup still matters in the young or atypical patient because you may need to deal with an AVM after the life-saving decompression.
Clinical course you must not miss
Early symptoms are posterior-fossa: occipital headache, vertigo, nausea, vomiting, truncal ataxia, dysarthria, and ipsilateral limb dysmetria. Gaze palsy, facial weakness, and a shrinking pupil signal brainstem compression. Level of consciousness may remain surprisingly good until it is not—deterioration often arrives over 12–24 hours as the hematoma expands or perihematomal edema peaks, but it can be abrupt. Once the patient is comatose with extensor posturing or apnea, the chance of a useful recovery after heroic decompression falls. The exam-ready rule: do not wait for coma to call the operating room.
Imaging thresholds
Noncontrast CT answers four questions: diameter and volume, fourth-ventricle patency, ambient and quadrigeminal cisterns, and hydrocephalus (temporal horns, third ventricle). Classic teaching used a maximum diameter >3 cm as a surgical trigger. A sphere 3 cm across is about 14 mL, which is why the 2022 AHA/ASA ICH guideline also lists cerebellar volume ≥15 mL alongside neurological deterioration, brainstem compression, and hydrocephalus from ventricular obstruction. Diameter and volume are complementary: an elongated 2.8 cm clot that effaces the fourth ventricle and compresses the pons is still an operation. A 3.2 cm peripheral hemispheric clot with open cisterns and a patent fourth ventricle may be watched in an ICU with a very low threshold to operate if the next CT or the next exam worsens.
| Finding | Why it matters | Usual action |
|---|---|---|
| Diameter >3 cm or volume ≥15 mL | Mass already large for the posterior fossa | Prompt evacuation, with or without EVD |
| Brainstem compression or cisternal effacement | Direct mechanical injury | Evacuate now; EVD is not a substitute |
| Fourth-ventricle obstruction or hydrocephalus | Rising intracranial pressure, falling consciousness | Evacuate; EVD is an adjunct |
| Neurological deterioration | Expansion or edema in a tight compartment | Do not wait for coma |
| Small hematoma, open fourth ventricle, normal exam | Compression not yet present | ICU medical care, serial CT, instant surgical backup |
Pontine (brainstem) hematoma is a different disease: primary surgical evacuation is generally not offered, and mixing “infratentorial” ICH score points with “operate on all infratentorial clots” is a common wrong answer.
The operation, and why EVD is not the operation
The standard procedure is suboccipital craniotomy or craniectomy with hematoma evacuation. Goals are to remove clot, decompress the posterior fossa, restore cerebrospinal fluid flow through the fourth ventricle, and inspect for an unexpected vascular malformation in selected patients. An EVD is often placed as an adjunct when hydrocephalus is present or expected, because it can rapidly lower supratentorial pressure and give you a drain if the aqueduct remains temporarily closed.
EVD alone is not enough if there is direct brainstem compression. The drain does not remove the cerebellar mass. If the fourth ventricle is obstructed, draining the lateral ventricles can create a pressure gradient that pulls the brainstem upward (upward transtentorial herniation). The 2022 AHA/ASA ICH guideline therefore lists EVD without hematoma removal as insufficient—and potentially harmful—when significant posterior-fossa mass effect is present (Class 3). A further practical trap: the pressure you read from a frontal EVD may not represent posterior-fossa pressure, so a “normal ICP number” does not reassure you that the pons is safe.
If hydrocephalus is the dominant problem and the hematoma is small without brainstem compression, an EVD can be a bridge—but those patients still belong in a center that can decompress in minutes if the exam changes. Do not park a large compressing cerebellar clot on a ward with an EVD and a plan to “see how they look in the morning.”
Distinct from hemispheric ICH surgery data
Carry a comparison table in your head, because vignettes will try to apply STICH II to the cerebellum.
| Feature | Hemispheric (STICH / STICH II / ENRICH) | Cerebellar hematoma |
|---|---|---|
| Landmark RCTs | STICH and STICH II: routine early open evacuation not generally beneficial; ENRICH: selected lobar MIPS | No equivalent RCT; posterior fossa excluded from STICH |
| Dominant injury | Primary tissue destruction plus expansion | Compression of brainstem and fourth ventricle |
| Size trigger | Volume 30–80 mL is an ENRICH window, not a craniotomy mandate | Diameter >3 cm or ~≥15 mL plus compression/hydrocephalus |
| Timing | Ultra-early open surgery can rebleed; ENRICH operated within 24 hours | Operate before coma; do not delay for a “STICH-style observation period” |
| EVD role | For IVH/hydrocephalus, not a substitute for ENRICH-style evacuation | Adjunct only; never the sole treatment of brainstem compression |
| Expected benefit | Functional outcome uncertain for open surgery | Mortality reduction with timely suboccipital decompression |
Medical ICH care still applies in the cerebellum: reverse anticoagulants the same way as in section 20.1, lower systolic pressure toward 140 mm Hg without ATACH-2 overshoot, control glucose and fever, and start pneumatic compression. Those measures do not replace the bone work.
Exam scenarios
- A 72-year-old on warfarin with a 4 cm cerebellar hematoma, drowsy, fourth ventricle gone: give 4-factor PCC plus vitamin K and take the patient to decompression; reversal without surgery leaves the mass in place.
- A GCS 15 patient with a 2 cm cerebellar clot, open fourth ventricle: ICU observation is reasonable, but the stem will often add vomiting and a later scan with hydrocephalus—that is now an operation, still before coma.
- An EVD is placed for hydrocephalus while the family “thinks about surgery” in a patient with brainstem compression: the correct critique is that the drain is not definitive therapy and may cause upward herniation.
- Someone quotes STICH II to argue that early surgery is unproven, so the cerebellar clot should be managed medically overnight: that trial does not apply.
If you remember only one sentence for 03.D.2: size >3 cm, brainstem compression, or hydrocephalus means evacuate now; do not wait for coma; an EVD is not a substitute.
A 70-year-old with a 3.4 cm cerebellar hematoma is drowsy, the fourth ventricle is effaced, and the ambient cisterns are tight. GCS is 12. Which plan is most appropriate?
In a patient with a large cerebellar hematoma causing direct brainstem compression, why is an EVD by itself the wrong definitive therapy?
A colleague argues that because STICH and STICH II were neutral, a 4 cm cerebellar hematoma with hydrocephalus should be managed medically unless the patient is already comatose. Which statement best corrects that reasoning?