Abugharib et al. examined the records of 657 men who had salvage radiation therapy (SRT) from 1986 to 2013 at the University of Michigan and the University of Texas Southwestern. They were all discovered to have detectable PSA following prostatectomy. Researchers were looking for evidence to confirm or contradict their hypothesis that earlier SRT had better outcomes.
They defined "earlier" in two ways:
1. At a lower PSA. Because of the treatment dates, there was relatively little data from ultrasensitive PSA tests. They divided PSA at the time of SRT into three categories:
- 0.01-0.2 ng/ml - the "very early salvage" cohort
- >0.2 - 0.5 ng/ml - the "early salvage" cohort
- >0.5 ng/ml - the "later salvage" cohort (0.5 was selected because the median PSA was 0.4 ng/ml)
- < 9 months
- 9-21 months
- 22-47 months
- > 48 months
- freedom from biochemical recurrence (PSA> 0.2 ng/ml) after SRT
- freedom from starting salvage, life-long androgen deprivation therapy (ADT) after SRT
- freedom from detectable metastases after SRT
- prostate cancer specific survival
After a median follow-up of 9.8 years, they found:
- The time in months since completion of prostatectomy had no bearing on any of the outcomes.
- The PSA at which they were treated has a major impact on all outcomes.
- The "early salvage" group had outcomes that were about twice as poor as those who had "very early salvage." This was true after correcting for all the variables (like Gleason score and positive margins) that would have made a difference.
- The "later salvage group" had outcomes that were about four times as poor as those who had "very early salvage." This was true after correcting for all the variables (like Gleason score and positive margins) that would have made a difference.
- 91% of the variance in biochemical recurrence after SRT was explained by the PSA at which patients were treated.
- Adjuvant ADT, which was given to 24% of patients for a median of 6 months (range 4-24 months), was significantly associated with freedom from biochemical recurrence after SRT. There were 40% fewer failures.
Researchers did not have data on PSA doubling time and velocity, and the number who had persistently elevated PSA, all of which almost certainly would affect outcomes. Perhaps such other variables as the length of the positive margins and the Gleason score there ought to be incorporated into a fuller analysis.
Patients who were treated at an early sign of detectable PSA (0.2-0.5) were twice as likely to develop metastases and die of prostate cancer as those who were treated at the earliest PSA (below 0.2). Those who waited for PSA to rise above 0.5 ng/ml were four times as likely to develop metastases and die from prostate cancer compared to those treated when PSA first became detectable.
We have three large randomized clinical trials proving that outcomes are diminished by about half by waiting rather than treating within the first 6 months, even before there are detectable PSAs (called adjuvant radiation). But few elect to have adjuvant radiation, and the number has been declining (see this link). To avoid overtreatment and protect patients from perhaps unnecessary side effects of SRT, early salvage has emerged as a compromise.
The authors point out that it may take 7 months or more for adequate healing of urinary and erectile complications (see this link). Also, this is an important decision for the patient, which he ought not make hastily. Yet here, more than in the primary therapy decision, very early action can save lives. As a compromise, they suggest early use of neoadjuvant ADT (prior to SRT) which could slow the cancer down and give tissues more time to heal. The extra time may help the patient recover better urinary function, if not erectile function.
"Our data would suggest potentially a traditional cut-off of 0.2 to define biochemical failure may be too late, and that at the first sign of a detectable PSA that SRT (or SRT + ADT) should be initiated."
This remains a difficult decision, and the patient with a detectable PSA after surgery should begin discussions with a good radiation oncologist as soon as possible. Age and comorbidities enter into the decision as well. Unfortunately at these low PSAs, even the most accurate of the new generation of PET scans are incapable of finding distant metastases that might help rule out those cases where SRT would be futile. Nomograms and Decipher scores may help in cases where the decision is equivocal.