How would you approach treatment of a recurrent brainstem metastasis following prior SRS?

Recurrent brainstem metastasis can be quite catastrophic. In the upfront setting, I recommend treating to an adequate tumoricidal dose (unfortunately I've seen folks back off significantly in the brainstem and this can result in lower LC rates). Having said that, should a recurrence occur, I would not treat with repeat SRS. There are many publications on repeat SRS for brain mets and while radionecrosis rates are not terribly high (~20-25%) [see: https://www.ncbi.nlm.nih.gov/pubmed/27494815 & https://www.ncbi.nlm.nih.gov/pubmed/29218301

that risk seems too high for brainstem lesions. If the patient has no other brain mets, or very few at the time of recurrence, one can consider partial brain irradiation to the brainstem lesion to 30 Gy in 10...

As @Dr. First Last has mentioned, recurrence/progression of an insufficiently treated metastasis in the brainstem is a very difficult predicament. I agree it is best to treat these lesions with a dose with a high probability of control. If the lesion is large enough that such a dose is worrisome in a single fraction, consider fSRS in 3 or 5fx. At UAB, we routinely treat intra-brainstem lesions too large for single-fx treatment to 25-30Gy/5fx with good control and minimal toxicity, which is a very conservative dose with respect to radionecrosis risk. (The conventionally computed BED of 25Gy/5fx is actually lower than 30Gy/10fx, yet the durable control rate is substantially higher (>80%) than WBRT (35-40%) (underscoring the failings of curren...