How do you approach CNS prophylaxis in patients with DLBCL?

Given the overall risk of CNS relapse for DLCL in our current era of chemoimmunotherapy is around 5%, the key to an algorithm for CNS prophylaxis is based on risk to avoid the risks of overtreatment. There are several methods of prognostication that have been published, which can identify groups with risk up to 10-15%. These are nicely outlined in the paper by Hall et al., PMID 29940062. Alternatively, we can use disease biology; for example, we know that "double hit" large cell lymphoma can have up to 25% risk primarily in the first two years of completion of therapy. These guidelines will still likely lead to over-treatment for some patients. The Wilson (Wilson et al., PMID 42274647) paper defines the conundrum of this question, given the...

I think the NCCN-CNS-IPI based on the German data is a reasonable place to start when it comes to making decisions regarding CNS prophylaxis. We typically do IT MTX for patients on the lower end of the risk spectrum and high-dose IV MTX for patients on the higher end of the risk spectrum.