How do you approach treatment of brain metastases of varying sizes with SRS?

The UAB institutional paradigm has evolved to the following current practice patterns.

Most patient's treatment plans are generated and delivered via HyperArc. Nearly all patients are treated on a Varian Edge with initial kv-kv, followed by CBCT/sim 6DOF alignment based on bony windowing with the VOI encompassing the skull but inferior border bisecting the mid-sagittal projection of the clivus (this mitigates mis-registration due to flexion at the atlanto-occipital joint. We monitor intrafraction motion with OSMS (now Humediq, formerly Vision/Align RT).

Small targets <1cm: 20Gy/1fx

Moderate targets 1-2.5/3cm: 18-20Gy/1fx

Larger targets 2.5/3-5cm: 27Gy/3fx or 30Gy/5fx

Largest targets >5cm: 24Gy/3fx or 25-27.5Gy/5fx

Targets in brainstem (pons or me...

So I think the question here is - would it be reasonable to consider treating a small brain metastasis that one would normally treat with single fraction SRS to fractionated SRS because there is plan to fractionate SRS a large brain metastasis?

I don't think there's a wrong answer here.

In general, I would prefer giving single fraction SRS to a small brain metastasis whether or not I plan to fractionate a large brain metastasis. That being said, if you look at the Alliance A071801 trial, smaller unresected brain metastases are required to receive fractionated SRS if patient were randomized to the fractionated arm.

No data to support this but I find that if I'm treating radio-resistant histology i.e., melanoma or RCC resection cavity and de novo mets, I like to give higher dose to de novo mets 22Gy in 1fx if small and fractionate large cavity 30Gy in 5 fractions if no gross tumor left behind. I agree, makes the planning not clean and harder to interpret brain dose but I feel 30/5 is too low a dose for de novo mets.