Design: Single-arm systematic review and meta-analysis
Population: 2,587 patients with brain metastases in 34 studies
Primary outcome: Survival, local control and radiation necrosis
Effect: 1-year local control 87% (95% CI 79–92); grade 3 radionecrosis 5% (3–8)

A single-arm meta-analysis pooled 34 studies with 2,587 patients with brain metastases, mainly from NSCLC and melanoma, treated with stereotactic radiosurgery or fractionated stereotactic radiotherapy alongside PD-1 or PD-L1 inhibitors.

One-year overall survival was 65%, local control 87% and distant intracranial progression-free survival about 40%. Grade 3 radiation necrosis occurred in 5% and grade 3 or worse neurological events in 7%, with no difference between melanoma and lung primaries.

Without a comparison group, this cannot say whether the combination adds benefit or increases necrosis compared with radiosurgery alone. It does provide a reference figure for consent: about one in twenty patients may develop significant radionecrosis. Distinguishing necrosis from progression on MRI remains a practical problem, and advanced imaging or multidisciplinary review often helps.

Quote about a 5% risk of grade 3 radionecrosis when consenting for radiosurgery during immunotherapy.
Plan MRI surveillance every two to three months after radiosurgery in patients on checkpoint inhibitors.
Review apparent progression at a neuro-oncology meeting before changing systemic therapy, as it may be necrosis.
Consider perfusion MRI or PET where available to separate necrosis from tumour.

Why it matters

It gives a usable figure for radionecrosis risk in a combination now common in practice.

The statistics, in plain English

Single-arm pooling gives an average rate across studies without a control group, so it describes outcomes but cannot measure the effect of adding immunotherapy. The wide interval for distant brain control (18% to 62%) shows how much studies varied.