Data shed light on salvage RT, immediate hormone Tx

August 15, 2004

I believe the evidence is accumulating in favor of earlier [androgen deprivation] therapy to provide a survival benefit.

The treatment of patients with advanced prostate cancer remains a challenge. Research presented at the 2004 AUA annual meeting did provide valuable insight into such key issues as the role of salvage radiation therapy and the benefit of immediate versus delayed androgen blockade. Five significant take-home points dealing with the treatment of advanced prostate cancer were presented at the meeting.

Pre-treatment PSA, PSA doubling time, Gleason grade, and seminal vesicle involvement are prognostic variables for estimating the success of salvage radiation therapy. Selected patients with high-grade disease, a short disease-free interval after prostatectomy, and/or rapid PSA doubling times-all poor risk factors-may still achieve a durable response to salvage radiation therapy if treated when PSA is <1.0 ng/mL.

A large, multicenter study of more than 500 patients reported that 3-year PSA progression-free probability was 89% for patients with Gleason 8-10; a history of positive surgical margins; and a PSA doubling time >10 months when radiation was given before the PSA level exceeded 2.0 ng/mL. Progression-free probability was 36% in the same circumstances when the PSA doubling time was £10 months. As with the majority of single-center studies published in the past several years, the PSA level at the time of salvage radiation delivery was an important factor in determining response durability, ie, the lower the PSA at initiation of external beam radiation therapy, the higher the likelihood of benefit.

Very few solid tumors that fail primary therapy yield to a salvage therapy with this degree of success. In the future, some prognostic indicator may be able to identify those patients with a rising PSA after radical prostatectomy who will absolutely fail salvage radiation, but that prognostic profile is not yet available.

I consider salvage radiation in all patients with a rising PSA after radical prostatectomy. The only exception is the rare patient whose PSA does not fall to an undetectable level after surgery.

Re-evaluation of data confirms that patients receiving immediate androgen blockade have superior survival compared with those who have deferred or delayed blockade. Cancer-specific survival at 11.9 years of follow-up is 85% for immediate blockade and 51% for delayed blockade.

The question of early androgen deprivation continues to challenge urologists. Is the timing of androgen deprivation related to a survival benefit? Can one achieve the same survival benefit when androgen deprivation is delivered later in the course of disease at the time of progression and, by doing so, avoid the side-effect profile associated with earlier and more prolonged androgen deprivation therapy? I believe the evidence is accumulating in favor of earlier therapy to provide a survival benefit.

The Eastern Cooperative Oncology Group (ECOG) EST3886 trial (N Engl J Med 1999; 341:1781) was updated to provide a longer follow-up and address the issue of potential grading imbalance between the two arms of the study. While only 50% of the slides of primary tumor histology could be retrieved, the central histologic re-evaluation of those that were available did not demonstrate an imbalance between the two arms and supported the integrity of the initial analysis. The survival benefit of early androgen deprivation was statistically confirmed.

A PSA doubling time of less than 10 months is a strong predictor of cause-specific survival after prostatectomy and can be used to define clinical trials for advanced prostate cancer.

Clinical trials studying prostate cancer desperately need a surrogate endpoint to expedite the evaluation of therapeutic options. Prolonged time to objective disease progression and death-the latter being the traditional FDA endpoint of efficacy-necessitates large trials of long duration.

PSA doubling time in several large database studies has demonstrated a strong correlation with subsequent disease progression and death. The cutpoint of PSA doubling time has varied from 3 months to 12 months depending upon the database analyzed.

A large Johns Hopkins database demonstrated a PSA doubling time £10 months to be a strong predictor of prostate cancer mortality, thereby identifying patients who require therapy and possibly for use as a surrogate for treatment effect. In June, the FDA held a workshop to investigate the possibility of employing surrogate endpoints. Their decisions and conclusions are awaited.