Skull Base & Orbit

Posterior Fossa Masses: Anchor to the Fourth Ventricle

Fellowship-level neuroradiology teaching by E. Brooke Schrickel, MD. Open this topic in the interactive reading room →

Watch · concise explainerVisual summary

Core clinical idea

In the small posterior-fossa box the differential is usually broad, so don't force a single diagnosis — anchor it. Fix a pediatric mass to the fourth ventricle (roof=medulloblastoma, floor=ependymoma, anterior=brainstem glioma, posterolateral=pilocytic), then let density, diffusion, and outlet-foramen extrusion narrow it, and treat the second finding (drop met, pial nodule, syndromic stigma) as part of the diagnosis.

Bottom line

In the pediatric posterior fossa, anchor the mass to the fourth ventricle, let density/diffusion/foraminal-extrusion narrow it, and image the spine when it seeds — a short honest differential beats a false slam dunk.

Core workstation questions

Where does this arise relative to the fourth ventricle — roof, floor, anterior, or posterolateral?
Is it dense/restricting (cellular tumor like medulloblastoma) or low-density/infiltrative?
Is it squeezing out the outlet foramina (Magendie/Luschka) — the ependymoma signature?
Does this tumor seed — do I need the whole-spine MRI before I'm done?
Is there a second finding (drop met, pial nodule, prominent vessels) that names a syndrome or changes the differential?
Is this overlapping fourth-ventricle mass one I can anchor, or do I owe a short ranked differential plus a neuraxis screen?
Does the history (corrected hyponatremia, alcoholic, transplant) obligate me to raise osmotic demyelination even if the scan is subtle or negative?
For a posterior-fossa hemorrhage, have I run SWI and decided whether follow-up or angiography is needed to find the cause?

What changes reporting / management

State the mass's origin relative to the fourth ventricle as the primary discriminator.
When calling medulloblastoma (or any CSF-seeding tumor), image the whole neuraxis — drop mets change staging and therapy.
In medulloblastoma follow-up, flag new restricted diffusion as a possible first sign of recurrence before enhancement returns.
For a pial hemangioblastoma nodule (especially multiple or young patient), recommend screening for VHL.
Use SWI/FLAIR to avoid scrolling past a small obex subependymoma.
For overlapping adult fourth-ventricle masses, give a short ranked differential and image the neuraxis when a seeding tumor is on the list.
When the story fits CPM and the pons looks subtly off, raise early osmotic demyelination and recommend short-interval follow-up — don't call it normal.
Treat DWI restriction as the earliest CPM sign; a negative early scan does not exclude it.
For a large posterior-fossa hemorrhage of unknown cause, run SWI and escalate to catheter angiography when a vascular malformation is plausible.
Run SWI on suspected cavernoma to find additional small lesions in the familial form.

Practical traps

Forcing a confident single diagnosis when an honest short ranked differential is correct.
Satisfaction of search — missing a second/third finding that actually makes the diagnosis.
Using the 'halo' to separate ependymoma from medulloblastoma — both can show it.
Calling a cyst-with-cobwebby-enhancement reflexively pilocytic instead of reopening the differential.
Scrolling past a sub-centimeter obex subependymoma because its signal is only slightly off CSF.
Collapsing a broad differential too early on overlapping fourth-ventricle masses.
Letting the clinical story rescue a diagnosis that doesn't declare on day one — and trusting a normal early scan in evolving disease.
Using adjacent flow voids to clinch choroid plexus papilloma (hemangioblastoma also shows them).
Accepting a benign popcorn appearance and closing the case without SWI or rebleed-risk consideration.
Reading a posterior-fossa hemorrhage as definitively infarct vs mass vs malformation acutely — they are indistinguishable early.

Teaching pearls

Origin relative to the fourth ventricle (roof/floor/anterior/posterolateral) places these tumors better than the enhancement pattern.
Toothpaste extrusion through Magendie and Luschka points to ependymoma; the halo doesn't separate it from medulloblastoma.
Medulloblastoma seeds — the spine MRI is part of the diagnosis, and new restriction can be the first sign of recurrence.
A sub-centimeter obex subependymoma hides on T2 — pick it up on FLAIR.
A pial hemangioblastoma nodule, especially multiple or young, means screen for VHL.
Flow voids next to a fourth-ventricle mass don't clinch choroid plexus papilloma — hemangioblastoma shows them too.
In CPM, DWI restriction is the earliest sign, the imaging lags the syndrome, and a normal early scan doesn't exclude it — make the diagnosis over time.
Osmotic demyelination happens with slow correction too, and even without hyponatremia in alcoholics and transplant patients.
A big posterior-fossa bleed hides its cause — run SWI and decide if follow-up or angiography is owed.
Benign popcorn = cavernoma, but always run SWI: the familial form is multiple and the rest may be tiny.

Teaching visuals

Posterior Fossa — medulloblastoma vs ependymoma (round ball vs plastic toothpaste)

Source lectures

Posterior Fossa (Part I)
Posterior Fossa (Part II)

Educational material for radiology residents and neuroradiology fellows. Nothing here drives individual patient care, and it contains no patient data.