Vestibular Schwannoma

1.Koos grade and modality choice

The Koos grading system stages tumor extent: I intracanalicular; II small extension into the cerebellopontine angle; III filling the CPA cistern and reaching the brainstem without compression; IV compressing/displacing the brainstem. Imaging uses thin-slice contrast T1 with heavily T2-weighted CISS/FIESTA for nerve and cisternal detail. Broad decision pattern:

• Observation: small, asymptomatic, or non-growing tumors, with serial MRI (many never grow).
• Microsurgery: large tumors (Koos IV), brainstem compression, significant mass effect or hydrocephalus, cystic/rapidly growing tumors, or younger patients preferring resection.
• Radiosurgery: small-to-moderate tumors (Koos I–III) that are growing or symptomatic; also for residual/recurrent tumor after surgery. NF2 is a special case requiring individualized, often hearing-prioritized strategy.

2.The dose de-escalation story

Early radiosurgery used marginal doses around 16 Gy and achieved excellent control but unacceptable cranial-neuropathy rates. Progressive de-escalation to 12–13 Gy (with some centers exploring 11 Gy) preserved tumor control while markedly reducing facial and trigeminal injury: one of radiosurgery's clearest dose-response/toxicity lessons. With a marginal dose of 12–13 Gy, contemporary series and the ISRS practice guideline report:

• Tumor control ~90–99% at 5 years (growth arrest or shrinkage; the goal is not disappearance).
• Facial nerve preservation ~95–100%.
• Trigeminal preservation ~79–99%.
• Serviceable hearing preservation ~41–79% early, declining over time; minimizing mean cochlear dose (< ~4 Gy) is the key modifiable factor.

A transient pseudoprogression (tumor swelling and possible central loss of enhancement) in the first 6–18 months is common and should not be mistaken for treatment failure; durable continued growth, not early enlargement, defines failure.

Tumor control and hearing preservation are different endpoints. Modern CNS guidance emphasizes that for intracanalicular or <2-cm tumors, SRS has not been shown superior to observation for preserving serviceable hearing. Hearing can decline over a decade despite durable tumor control, so counseling should not imply that treating a small stable tumor protects hearing. Better baseline hearing, smaller tumors, a marginal dose <13 Gy, and lower cochlear dose are favorable features, not guarantees.

Technically, the marginal dose is prescribed to a steep isodose: classically the 50% isodose line with Gamma Knife and the ~80% line with LINAC-based systems, and doses above 13 Gy measurably increase facial paralysis, trigeminal dysfunction, and hearing loss without improving control. Long-term series show local control of roughly 90–94% at 10 years, modestly below the 5-year figure, while serviceable hearing erodes with time regardless of modality, often ~50–70% preserved at 3–5 years but falling toward ~30–50% by 10 years. Hearing preservation is most likely with a small tumor (< ~1.5–2 cm), Koos I–II, good baseline (Gardner-Robertson) hearing, short symptom duration, and a low mean cochlear dose.

Two practical nuances on hearing. First, the cochlear constraint: for single-fraction SRS the working target is a mean cochlear dose at or below 4 Gy, with doses above 4.2 Gy carrying higher risk and TG-101 constraining cochlear point maximum to 9 Gy. Do not interchange mean and maximum thresholds. For fractionated SRT, a cochlear dose below 35 Gy is associated with hearing preservation. Cochlear sparing should not be pursued at the expense of tumor coverage or control. Second, better baseline hearing, Gardner-Robertson class I, shorter symptom duration, smaller tumor, marginal dose at or below 13 Gy, and low mean cochlear dose predict better hearing outcomes. Those are predictors within a treated cohort, not an argument for treating earlier to protect hearing: neither randomized nor matched comparisons have shown a hearing advantage for upfront radiosurgery over observation. One propensity-matched single-institution series reported better long-term hearing preservation prescribing to the 40% rather than the 50% isodose line, at 83.3% versus 57.1% at 5 years, but it is retrospective, single-center, and unreplicated. On platform, Gamma Knife plans are characteristically inhomogeneous with a steep dose gradient and robotic-LINAC plans more homogeneous; the 2025 CNS guideline found no studies comparing Gamma Knife, LINAC, and proton radiosurgery against one another and therefore made no recommendation on outcome by platform. Treat platform as a matter of local expertise and plan quality rather than a decision with comparative evidence behind it.

3.SRS vs microsurgery vs observation; larger tumors; NF2

For small-to-moderate tumors, SRS and microsurgery achieve comparable long-term control, with SRS offering lower immediate morbidity and microsurgery offering immediate debulking and tissue diagnosis; the choice is individualized. Larger tumors approaching or exceeding the single-fraction comfort zone may be treated with hypofractionated SRT (e.g., 3–5 fractions) to respect brainstem and nerve tolerance, or resected. Hydrocephalus or significant brainstem compression generally favors surgery first. In NF2-related schwannomatosis, formerly neurofibromatosis type 2, bilateral disease and the need to preserve hearing as long as possible make management highly individualized, and it often integrates radiosurgery, surgery, and systemic therapy. Bevacizumab remains the best-established systemic agent for hearing response, and brigatinib showed radiographic and hearing activity in the INTUITT-NF2 platform trial.

Prospective comparative cohorts in small tumors found better facial-nerve and early hearing outcomes after radiosurgery than microsurgery, but treatment allocation was not randomized and long-term hearing declines in both groups. Three caveats remain important: cystic tumors may enlarge unpredictably; malignant transformation after SRS is exceedingly rare and causality is difficult to establish; and salvage surgery after true post-SRS progression can be technically more difficult: a point to weigh in young patients with very long horizons.

4.The comparative evidence, and the hearing question

Vestibular schwannoma now has two randomized trials alongside a large body of prospective comparative cohorts, and several genuinely live debates about when and how to treat. V-REX randomized 100 patients with newly diagnosed unilateral tumors under 2 cm to upfront Gamma Knife radiosurgery or wait-and-scan. At 4 years the geometric mean tumor volume ratio was 0.87 with radiosurgery versus 1.51 with observation, and 6% of the radiosurgery group versus 42% of the observation group required additional treatment, with no difference in hearing, quality of life, or symptom burden. ACOUNEU randomized patients with serviceable hearing to 18 Gy in 3 fractions or single-session radiosurgery and found no difference in hearing preservation at 36 months, with 92% local control at a median 62 months.

Open controversies:

• Observation versus upfront SRS. For a small, minimally symptomatic tumor, both are defensible. Current CNS guidance does not find SRS superior to observation for hearing preservation in intracanalicular or <2-cm tumors, so documented growth, symptoms, age, and patient preference should drive treatment.
• The cochlear dose threshold. Hearing preservation tracks dose to the cochlea, but the precise limit is debated: a mean cochlear dose around 4 Gy (some argue lower, ~3 Gy) is the working target, with marginal dose kept near 12–13 Gy.
• Single-fraction versus fractionated. Some centers favor fractionated SRT for larger or hearing-priority tumors, but ACOUNEU found no difference in hearing preservation at 36 months between 18 Gy in 3 fractions and single-session SRS. Combined with the CNS recommendation favoring single fraction on cranial-nerve grounds, the case for fractionating a small tumor to protect hearing is weak.
• Large and cystic tumors. Bulky tumors with brainstem compression and rapidly enlarging cystic lesions remain surgical; the boundary of safe SRS by size is a matter of judgment.