The immortal debate.
Top Line: With ultra-sensitive PSA tests widely available for close surveillance following prostatectomy, many people have landed on an “early salvage” approach as a solution to the decades long adjuvant-versus-salvage radiation debate.
The Study: That was fun while it lasted, but the gang’s back together again to reignite the debate. These prolific authors suggest, particularly when looking at very high risk pathology, that immortal time bias may be the only driver of small reported reductions in time to progression with the early salvage versus adjuvant approach. Now stay with us. Picture a man with a PSA rise to a level that would be considered progression (let’s say 0.4) over the 6 month period following surgery whose PSA wouldn’t have responded to either aRT or sRT. On the adjuvant arm, he would receive radiation post-op and be marked as progression at his first post-radiation lab check at roughly 6 months post-op. On the salvage arm, he would receive radiation within 4 months of his trigger PSA ≥ 0.2 (let’s say at 2 months post-op) and not have another lab check until as long as 3 months after his 2 months of radiation is delivered, meaning progression could be recorded as long as 10 months post-op. So the same man with the same disease course has “achieved’ almost double the time to progression with salvage radiation. Well that ain’t right. With that in mind, here's a multi-institutional retrospective analysis that includes 1383 men with post-prostatectomy adverse pathology (read: pN1, ECE and Gleason 8+, or invasion into seminal vesicles or adjacent organs).who received adjuvant (n=428) or early salvage (n=965) radiation. The kicker is, they excluded all men like the one above who never saw a decrease in PSA, i.e. the men who didn’t benefit from either approach. Alas, after propensity matching for systemic therapy, etc, all-cause mortality was significantly reduced among men receiving adjuvant versus early salvage radiation (HR 0.66), particularly when node-positive disease was excluded (HR 0.33).
TBL: Recognizing men with adverse pathology comprised at most 9%-17% of the enrollees in the three randomized trials designed to answer this question, we hope someone will still attempt a pooled analysis excluding “nonresponders.” | Tilki, J Clin Oncol 2021