20.3 Intraventricular Hemorrhage (03.D.3)

Intraventricular hemorrhage (IVH) is blood in the ventricular system. Most adult neurocritical-care IVH is secondary—a parenchymal hematoma (often thalamic or caudate) ruptures into the ventricle. Primary IVH, with little or no parenchymal clot, is less common and should trigger a hunt for vascular malformation, aneurysm, or choroid-plexus source. Either way, blood that fills the third or fourth ventricle turns a hemispheric problem into an obstructive hydrocephalus problem. Blueprint item 03.D.3 asks you to grade the burden, know when an EVD is required, and not over-read CLEAR III as a positive functional-outcome trial.

How IVH injures the brain

Clot in the ventricles does three things. First, it blocks cerebrospinal fluid at the foramen of Monro, aqueduct, or fourth-ventricle outlets, producing acute hydrocephalus and a falling GCS. Second, the mass of a casted lateral ventricle can itself shift midline structures. Third, blood-breakdown products inflame ependyma and may contribute later to communicating hydrocephalus and shunt dependence. IVH is an independent predictor of death in ICH cohorts and is worth 1 point on the ICH score even before you quantify how much blood is present. Patients with IVH extension have substantially higher mortality than ICH without ventricular blood in observational series; among those with IVH, heavier Graeb burdens track worse outcomes.

Graeb and modified Graeb, conceptually

You do not need to memorize every cell of a research scale, but you must know what the numbers mean.

The original Graeb score (0–12) assigns up to 4 points to each lateral ventricle (trace blood; less than half filled; more than half filled without expansion; completely filled and expanded) and up to 2 points each to the third and fourth ventricles (blood present versus filled and expanded). A score of 0 is no IVH. Scores in the upper half of the scale imply a large clot burden and a high chance of obstruction.

The modified Graeb score (mGS) expands the same idea to a 0–32 range by scoring additional compartments—each lateral body, each occipital horn, each temporal horn, plus third and fourth ventricles—and adding a point when a compartment is expanded because of clot. Modified Graeb correlates more tightly with measured IVH volume and with outcome than the original 12-point score, which is why modern trials (including CLEAR III imaging cores) often report it. Exam use: a high Graeb or modified Graeb with a closed third or fourth ventricle is the patient who needs CSF diversion, not a patient you observe because “IVH is just one ICH-score point.”

Casted ventricles means a compartment is packed with clot rather than layered blood-CSF levels. Casts at the foramen of Monro can trap a lateral ventricle (it dilates in isolation). Casts of the third and fourth ventricles produce the classic acute obstructive picture. Trapping and complete casts are the anatomic reasons a second EVD is considered: one catheter in cerebrospinal fluid for ICP control, another in the dominant cast if you are attempting clearance.

EVD for obstructive hydrocephalus

The 2022 AHA/ASA ICH guideline supports ventricular drainage when hydrocephalus contributes to reduced consciousness, to reduce mortality. The ideal story is a patient who was speaking and then becomes sleepy as temporal horns enlarge—placing an EVD can reverse that decline in minutes. EVD is not a treatment of the underlying putaminal destruction, and survival after drainage is often at the cost of severe disability when the parenchymal injury is already devastating. That does not make the drain optional when the immediate threat is herniation from hydrocephalus.

Practical points the exam likes:

• Place the catheter using a standard frontal landmark (Kocher’s point is the usual adult entry) and confirm position on CT before you inject anything.
• Level the transducer at the external auditory canal (or the institution’s tragus convention) and document whether the drain is open at a set height or cycling.
• A stability CT after placement is mandatory before intraventricular drug instillation; catheter-tract hemorrhage is common enough that CLEAR III required radiographic stability.
• Wean by raising the drain or clamping with repeat imaging and a clinical exam; failed weans imply communicating hydrocephalus and possible shunt.
• In cerebellar ICH, remember section 20.2: an EVD for hydrocephalus does not replace posterior-fossa decompression when the brainstem is compressed.

CLEAR III: faster clearance, not a proven outcome therapy

CLEAR III (Hanley et al., Lancet 2017) randomized 500 patients who already had an EVD for IVH obstructing the third or fourth ventricle, with a stable parenchymal ICH <30 mL, to alteplase 1 mg every 8 hours through the drain (up to 12 doses) versus saline. Dosing stopped when the third and fourth ventricles opened, about 80% of IVH was gone, 12 doses were given, or a safety endpoint occurred. Randomization was relatively late (median about 52 hours from onset).

Alteplase opened the third and fourth ventricles faster (p < 0.0001). The primary endpoint—modified Rankin Scale 0–3 at 180 days—was 48% versus 45% (risk ratio 1.06, 95% CI 0.88–1.28, p = 0.554). That is a neutral functional result. Secondary analyses found lower mortality with alteplase (adjusted odds ratio for death about 0.50) but more survivors with severe disability, which is why commentators describe a shift from death into mRS 4–5 rather than a proven independence benefit. Bacterial ventriculitis was 7% versus 12% (favoring alteplase in that comparison); symptomatic bleeding was similar (about 2% in each arm). Greater percent clot removal associated with better mRS in post hoc analyses, and dual catheters or a catheter sitting in the dominant cast cleared more blood—but those findings are not the primary endpoint.

Do not teach IR-tPA through an EVD as proven functional-outcome therapy. You may discuss it as a protocol that clears blood and appeared safe in a highly selected EVD population, with a mortality signal that did not convert into the trial’s pre-specified good-outcome win. Current practice should not treat CLEAR III as a mandate to inject alteplase into every IVH.

Casted ventricles, dual drains, and infection

When both lateral ventricles are packed, a single contralateral catheter sitting in CSF may control ICP and even open the midline ventricles with saline or alteplase, yet leave a cast in the opposite body. CLEAR III surgical guidance suggested a second catheter for trapping, casting, or mass effect. Dual EVDs increase nursing complexity and the number of access points for bacteria. They are a response to anatomy, not a routine for trace occipital-horn blood.

Infection is the complication you will be asked to prevent. Risk rises with dwell time, frequent CSF sampling, disconnected stopcocks, and CSF leak around the tunnel. Use a closed system, tunnel the catheter, avoid “just-in-case” daily cultures in a well patient, and remove the drain when hydrocephalus physiology is gone. Antibiotic-impregnated catheters are used in many units; they do not make sloppy technique safe. Catheter-tract hemorrhage, overdrainage with hygromas, and upward herniation in posterior-fossa mass lesions complete the harm list.

Primary IVH, young patients, or IVH without hypertension still need vascular imaging. Do not attribute a pure ventricular flood in a 34-year-old to “hypertensive ICH” and skip CTA or catheter angiography.

Exam traps: calling CLEAR III a positive mRS trial; injecting alteplase through an EVD before a stability CT; treating every IVH with a drain even when the ventricles are not obstructed and the patient is alert; using a frontal EVD as the only treatment of a compressing cerebellar hematoma; forgetting that Graeb 1 (trace occipital blood) and a complete ventricular cast are not the same disease.