Do you place any constraint on Dmax when developing plans for linac-based SRS or SBRT?

Very rarely. I was a gamma knife SRS person before I started linac-based SRS so I am very comfortable with >120% hot spots. For brain linac based SRS we regularly push hotspots to 150% to get the steepest dose falloff (most similar to gamma knife plans prescribing to the 50% isodose curve). For body you don't usually see more than 120% without specifically asking but if I am convinced the GTV is tumor ONLY (no normal tissue infiltrating within the GTV), I don't hesitate instructing dosimetry to ...

This is the number one thing I see leading to poor quality in linac SRS and SBRT plans, especially when reviewing plans from dosimetrists and physicists who trained on linacs in the era before linac SRS and SBRT became more common.

Merely the act of asking in your cost function for homogeneity in an SRS/SBRT plan will worsen your dose falloff at virtually all clinically meaningful dose levels.

There is simply no reason to penalize a hotspot within tumor. And if your facility does not have the ab...

Agree with the hotspot notion. First, we have long history of knowing that it's safe (particularly without intervening OAR). Second, it may well be a driver of control! See our publication on this. Hotspots are a natural consequence of creating highly conformal 3D plans, where the collimator or block is right at the tumor margin (or even negative block margins with lung SBRT!), requiring increased fluence to get coverage.

As we've moved into an IMRT/VMAT era, the inverse planning algorithms intr...