Patterns of Failure Following the Brachytherapy Management of Prostate Cancer

Reviewer: Christopher Dolinsky, MD
University of Pennsylvania School of Medicine
Last Modified: November 8, 2006

Presenter: R.G. Stock
Presenter's Affiliation: Mount Sinai Medial Center, New York, NY
Type of Session: Scientific

Background

There are relatively few studies with long-term outcomes for men who have been treated with radioactive seed implant (brachytherapy) for prostate cancer.
It is important to understand patterns of failure after brachytherapy in order to decide which patients should be offered this treatment modality.

Materials and Methods

Between 1990 and 2006, 2,456 patients with clinical stage T1-T3 prostate cancers were treated with brachytherapy at Mount Sinai Medical Center.
54% of patients were considered low risk, 22% were considered intermediate risk, and 33% were considered high risk, based on the D’Amico risk group stratification.
Median follow-up was 82 months (range 24 to 184 months).
Biochemical failure was defined using the ASTRO definition (3 consecutive rises in PSA).

Results

184 patients developed PSA failure.
104 of the 184 PSA failures had a 2-year post treatment biopsy.
Amongst the patients who had PSA failure, 22% were low risk, 14% were intermediate risk, and 64% were high risk.
Amongst the patients with PSA failure, the PSA doubling time was <3 months in 21% of the patients, 3-6 months in 17%, 6-10 months in 16%, and >10 months in 46%.
When looking at a patient’s likelihood of having a positive post-treatment biopsy using PSA doubling time, there were zero patients with doubling times of either <3 months or 3-6 months who had a positive biopsy, but 30% had a doubling time from 6-10 months, and 49% had a doubling time >10 months.
When comparing the biologically effective dose received for patients with a positive biopsy (either >= 150Gy or < 150Gy), 44% received <150Gy and 24% received >=150Gy.
37 patients developed distant metastases.
14 of these 37 underwent post-treatment biopsy, and all biopsies were negative.
The 10-year actuarial freedom from developing distant metastases was 76%.
Gleason score was significantly related to freedom from developing distant metastases (89% for Gleason <=6, 75% for Gleason 7, and 50% for Gleason 8-10, p<0.001).
The amount of time that elapsed between brachytherapy and PSA failure was related to freedom from developing distant metastases (96% for >3 years, 78% for 2 to 3 years, 65% for 1 to 2 years, and 58% for less than 1 year, p=0.0001).
PSA doubling time was significantly related to freedom from developing distant metastases (92% for >10 months, 84.5% for 6 to 10 months, 66% for 3 to 6 months, and 22% for <3 months, p<0.0001).

Author's Conclusions

High risk patients make up the majority of treatment failures.
Local failure occurred in one-third of patients, and was associated with lower radiation doses and longer doubling times.
Distant metastasis occurred in one-fourth of patients, and was associated with shorter doubling times, early PSA failure, and higher Gleason scores.

Clinical/Scientific Implications

The authors present a retrospective look at a large number of patients who underwent prostate seed implant at one institution. It is important to continue to update and publish the experience of single institutions, particularly when a novel treatment strategy is employed. These data suggest that seed implant should probably be reserved for low risk patients and highly selected intermediate risk patients. The risk of relapse appears too high to justify seed implant in high risk patients.

It would be nice to have an analysis of the toxicities related to these implants. It may be the case that external beam radiation can produce similar and/or superior outcomes with less acute and long-term toxicities than brachytherapy. However, it will be hard to compare them head-to-head, because the quality of an implant is very operator-dependent. Centers with a high volume of cases and experience with brachytherapy tend to have superior outcomes when compared to centers that perform them sporadically.