What dose constraints would you use if reirradiating the prostate using SBRT?

This is a difficult question to answer in this forum due to the limited availability of prospective data on which to define optimal OAR tolerance. The answer would depend on the dose, fractionation, treatment volume (GTV vs. whole gland), and technique which would be employed (CK vs. IMRT/VMAT). There are multiple other issues as well which arise such as how to delineate the GTV if that is your intended target volume (MRI vs. PET) or dose homogeneity if delivering treatment to the entire gland (homogeneous as in D'Agostino et al., PMID 30825495 vs. heterogeneous as in Fuller et al., PMID 26059509).

There was a recent attempt by ESTRO/ACROP to arrive at a consensus for dose constraints in this situation, however, they failed to do so. The authors were split on whether or not the dose constraints used should depend on the interval between prior RT and salvage as well as the actual dose constraints themselves (see supplement). The results of NCT00851916 have been published, however, in this trial, they employed treatment to the entire gland which is contrary to the ESTRO/ACROP consensus recommendations. Additionally, interim results of NCT03073278, a dose-escalation trial, have been published as well. It’s notable that their most significant reported toxicity was from the extravasation of hydrogel spacer in one patient as the placement can be technically challenging in these cases. Further, there is an ongoing French trial, GETUG-AGU 31 (NCT03438552) with dose constraints based on a prior multi-institutional retrospective review by GETUG. At our center, we are also currently actively recruiting patients for NCT03253744, a phase I dose-escalation trial, with treatment directed to the GTV as identified by [18F] DCFPyL-PSMA PET/CT. 

Treatment in this setting can produce numerous technical challenges, and the current NCCN guidelines do not recommend SBRT as a salvage option for local failure. They instead recommend salvage brachytherapy (BT) which has much more data validating its use. My appraisal of the available data does lead me to believe that salvage SBRT will eventually win out over salvage BT owing to a possibly better therapeutic index to salvage BT with similar efficacy, decreased toxicity, and less need for procedural expertise. For example, RTOG 0526, a salvage brachytherapy study had a rate of grade 3 toxicity of ~ 15% while other studies show slightly higher rates of ≥G3 toxicity (López et al., PMID 31570236, Smith et al., PMID 33075391) for salvage BT. Grade 3 events seem to be less frequent in the available reports of salvage SBRT, although there has been no direct comparison. A recent meta-analysis of salvage options for patients with radiorecurrent prostate cancer summarizes this nicely with similar point estimates of efficacy and reduced toxicity for salvage SBRT.

For now, going back to the dose constraints, if you are very motivated, the citations above of the available trial reports do list their dose constraints, although, in my opinion, this type of treatment is currently best delivered on clinical trial at this time or at the least, a high volume center with substantial experience.