PSA "Bounce" After Radiation for Prostate Cancer Does Not Appear Related to Survival, Cure

Jill Taylor

October 07, 2004

Oct. 7, 2005 (Atlanta) — A posttreatment "bounce" in prostate-specific antigen (PSA) levels does not appear to indicate that patients will die of prostate cancer any sooner than patients without a bounce, new research shows.

Although the PSA bounce phenomenon has been documented to occur in 20% of patients or more treated with radiation therapy by either external beam or implants, lead investigator Eric M. Horwitz, MD, from Fox Chase Cancer Center in Philadelphia, Pennsylvania, said that it is not commonly recognized.

"The phenomenon was first recognized where one to three years after treatment, men who had implants experienced a rise and fall in PSA," Dr. Horwitz told Medscape. "After treatment, the PSA should go down and stay relatively level, so at first people thought that treatment hadn't worked."

Futhermore, the same phenomenon was noted to occur at the same frequency in prostate cancer patients treated by external beam radiation therapy (EBRT). While previous studies have shown no clinical significance for patients with a PSA bounce compared with nonbounce (NB) patients, Dr. Horwitz pointed out that there was no investigation into whether the PSA bounce had implications for long-term cancer cure.

"We found that there is a difference in biochemical control, but it did not translate into clinical failure. Men who experienced a PSA bounce did not have more distant metastases or local failures, and they did not die of prostate cancer more often than men who did not have a PSA bounce," Dr. Horwitz said.

To determine if there was a difference in biochemical and clinical control between the bounce and NB patients, the researchers analyzed pooled data from nine institutions of 4,839 patients with T1-T2 prostate cancer treated with EBRT alone.

Patients were treated with EBRT (>60 Gy) without hormone therapy and received follow-up for a median of 6.3 years. For study purposes, a posttreatment PSA bounce was defined as an increase of 0.2 ng/mL or more over a previous PSA measurement, followed by a decline. Study end points included bNED failure (BF), distant failure (DF), cause-specific failure (CSF), and overall survival (OS).

For analysis, patients were stratified by pretreatment PSA level, Gleason score (GS), T stage, age, EBRT dose, and risk group (low risk: T1/2a, GS <= 6, and PSA <= 10 ng/mL; intermediate risk: T1/2a, GS = 7, and PSA = 10-20 ng/mL, or T2b/c, GS = 7, and PSA = 20 ng/mL; high risk: GS = 8-10 or PSA > 20 ng/mL). Univariate analysis (UVA), multivariate analysis (MVA), and a landmark analysis between PSA bounce patients and NB patients were performed for patients completing three or more years of follow-up with no clinical failure (CF) or BF during that time.

Of the study population, 978 patients were observed to have at least one posttreatment PSA bounce. For three subgroups (risk group, pretreatment PSA, and age), UVA analysis showed significant differences in bounce-free outcome between the bounce and NB patients. At five years, high-risk patients or patients with pretreatment PSA higher than 20 ng/mL had a probability of 67% and 63%, respectively, of remaining bounce-free compared with low-risk patients and those with PSA levels less than 10 ng/mL: 76% ( P < .0001) and 78% ( P = .0008), respectively; patients younger than 70 years had a 72% probability of remaining bounce-free compared with 75% for older patients ( P = .04).

In the landmark analysis, NB patients had 72% bNED control at 10 years compared with 58% for bounce patients with a hazard ratio (HR) of 1.73 ( P < .0001). UVA of the subsets demonstrated that the effect of bounce on subsequent BF was significant across all the groups except those with a GS = 8-10.

The effect of a bounce remained significant on MVA when controlling for the effect of the other covariates (HR, 1.67; P < .0001). No significant difference in DF, CSF, or OS was observed, and although the median height of the bounce was greater for patients receiving less than 70 Gy (0.8 ng/mL) compared with those receiving 70 Gy or higher (0.7 ng/mL), this did not translate into significant differences in BF.

"These results are clinically important in that if people experience a bounce, we can reassure them that there is no long-term difference in survival or cure," Dr. Horwitz told Medscape. "Also, they don't necessarily have to get hormone treatment, which is the established treatment. Before people realized what the bounce meant, patients would be given treatment with hormones and be exposed to those side effects."

The study was independently funded. The lead investigator reports no financial conflicts of interest.

ASTRO 46th Annual Meeting: Abstract 171. Presented Oct. 6, 2004.

Reviewed by Gary D. Vogin, MD


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.