Long-term outcomes after treatment with external beam radiation therapy and palladium 103 for patients with higher risk prostate carcinoma

Influence of prostatic acid phosphatase

Authors


Abstract

BACKGROUND

The objective of this study was to define the long-term prognostic significance of prostatic acid phosphatase (PAP) levels in patients with higher risk, early-stage prostate carcinoma.

METHODS

One hundred sixty-one consecutive patients with Stage T1–T3 prostate carcinoma (according to the 1992 criteria of the American Joint Committee on Cancer) were treated from 1992 through 1996. Each patient had a Gleason score ≥ 7 and/or a prostate specific antigen (PSA) level > 10 ng/mL. The original biopsy slides for 130 of 161 patients were retrieved and rereviewed by a single pathologist (L.T.). Enzymatic PAP measurements were determined using a standard method. Values up to 2.5 Units were considered normal. Patients received 41 grays (Gy) of external beam radiation therapy to a limited pelvic field followed 4 weeks later by a palladium 103 (Pd-103) boost using transrectal ultrasound and fluoroscopic guidance as described previously. The prescribed minimum Pd-103 dose to the prostate was 80 Gy (pre-National Institute of Standards and Technology [NIST]-99). Freedom from biochemical failure was defined as a serum PSA level ≤0.2 ng/mL at last follow-up.

RESULTS

There was little correlation between pretreatment PSA levels, Gleason scores, and PAP measurements. Thirty-eight patients developed biochemical failure. The overall actuarial freedom from biochemical progression at 10 years is 79%, with 118 patients followed for > 5 years. In a multivariate Cox proportional hazards analysis that considered each factor as a continuous variable, the strongest predictor of failure was PAP (P = 0.0001), followed by Gleason score (P = 0.13), and PSA (P = 0.04). PAP was especially helpful in stratifying patients with pretreatment PSA levels between 4 ng/mL and 20 ng/mL, for whom the prognosis does not different when they are subdivided into PSA categories. When the PAP subgroup analysis was limited to this relatively favorable group, there was a wide range of prognoses.

CONCLUSIONS

The biochemical cure rate was remarkably high among the 161 patients evaluated. The fact that the PAP was the strongest predictor of long-term biochemical failure in patients with otherwise higher risk features reported here suggests that it may be a more accurate indicator of micrometastatic disease compared with the Gleason score and the PSA level. This report adds to the rationale for reintroducing PAP measurement into general practice. Cancer 2003;97:979–83. © 2003 American Cancer Society.

DOI 10.1002/cncr.11154

Long before prostate specific antigen (PSA) was discovered, prostatic acid phosphatase (PAP) was the principle prognostic biochemical parameter for patients with prostatic carcinoma.1 Because of its greater sensitivity for detecting disease and monitoring residual prostate carcinoma, PSA measurement rapidly replaced PAP measurement in the early 1990s as the principle biochemical parameter for the management of patients with prostate carcinoma.2 More recently, however, several investigators have shown that PAP measurement offers prognostic information in addition to disease stage, PSA level, and Gleason score, and they have suggested reintroducing PAP measurement into clinical practice.3, 4

For patients who were treated with radiation therapy before the introduction of high-dose conformal beam radiation therapy or good-quality implants, a pretreatment PSA level > 10 ng/nL or < 10 ng/mL was a strong stratification factor.5, 6 However, with higher beam or brachytherapy doses, the prognoses appear remarkably uniform for patients with pretreatment PSA levels between 0 ng/mL and 20 ng/mL or Gleason scores of 2–7.7 We previously reported some prognostic significance of pretreatment PAP values for patients undergoing brachytherapy.8 With longer follow-up, PAP values appear to achieve increasing significance over PSA levels or Gleason scores in predicting prognosis. It may be especially valuable in refining the prognostic categorization of patients within groups with PSA levels of 0–20 ng/mL or Gleason scores of 5–7. Accordingly, we report here the long-term prognostic significance of PAP measurement in patients with higher risk, early-stage prostate carcinoma.

MATERIALS AND METHODS

One hundred sixty-one consecutive patients with Stage T1–T3 prostate carcinoma (according to the 1992 criteria of the American Joint Committee on Cancer9) were treated from 1992 through 1996. Each patient had a Gleason score ≥ 7 and/or a PSA level > 10 ng/mL (Fig. 1). To reduce subjectivity, the clinical stage was not included in this risk stratification. Patient ages ranged from 45 years to 88 years (median, 67 years). Seventy-eight patients met both criteria. All patients who met one criterion or both criteria were encouraged strongly to have palladium 103 (Pd-103) plus supplemental external beam radiation therapy (EBRT): There was no intention to treat patients with more advanced disease with EBRT alone.

Figure 1.

Patient characteristics for 161 patients who had pretreatment prostate specific antigen (PSA) levels > 10 ng/mL or Gleason scores ≥ 7 and received treatment with palladium 103 plus 41-gray external beam radiation therapy.

Due to concern regarding interobserver variability in grading, the original biopsy slides for 130 of 161 patients were retrieved and rereviewed by a single pathologist (L.T.) at the University of Washington to verify independently the patients' tumor grade. Patients for whom biopsy slides could not be reviewed were not included in the statistical analysis of Gleason score and biochemical outcomes.

Only one patient underwent a staging pelvic lymphadenectomy. Postimplantation biopsies were not performed routinely. Enzymatic PAP values were determined by using the method of Roy and colleagues.10 Values up to 2.5 U were considered normal.

Patients received 41 grays (Gy) of EBRT to a limited pelvic field followed 4 weeks later by a Pd-103 boost using transrectal ultrasound and fluoroscopic guidance as described previously.11 The prescribed minimum Pd-103 dose to the prostate was 80 Gy (pre-NIST-99). Treatment planning parameters have been described previously.11 A median of 89 mCi Pd-103 was implanted with a range of 51–141 mCi. The median source strength was 1.4 mCi (range,1.0–1.6 mCi/source), and the median number of sources was 64 (range, 36–101 sources). Patients generally were followed at 3 months, 6 months, and 12 months and every 6–12 months thereafter.

Freedom from biochemical failure was defined as a serum PSA level ≤ 0.2 ng/mL at last follow-up. Patients were censored at last follow-up if their serum PSA level still was decreasing (one patient). Patients with serum PSA levels that nadired at > 0.2 ng/mL were scored as failures at the time at that their PSA nadired. The follow-up period for nonfailing patients ranged from 0.5 years to 10.0 years (median, 7.0 years). Freedom from failure curves were calculated by the method of Kaplan and Meier.12 Differences between groups were determined by the log-rank method.

RESULTS

There was little correlation between pretreatment PSA levels, Gleason scores, or PAP levels (Fig. 2). Thirty-eight patients developed biochemical failure. The overall actuarial rate of freedom from biochemical progression at 10 years is 79%, with 118 patients who were followed for > 5 years (Fig. 3). Follow-up prostate biopsies were performed on all failing patients. There were no pathologically documented local failures, nor was there clinical evidence of local failure. Forty-eight patients received a short course of pretreatment androgen ablation therapy, which did not affect the failure rates (P = 0.28).

Figure 2.

Correlations between prostatic acid phosphatase (PAP) measurement (in Units) and prostate specific antigen (PSA) level (in ng/mL) or Gleason score (r: correlation coefficient).

Figure 3.

Freedom from biochemical progression for 161 patients (pts) with prostate specific antigen (PSA) levels > 10 ng/mL or Gleason scores ≥ 7 who were treated with palladium 103 (Pd-103) plus 41-gray external beam radiation therapy (EBRT).

Progressively more elevated pretreatment PSA levels, Gleason scores, and PAP levels each were associated with a progressively higher failure rate (Fig. 4). The presence of an elevated PAP level was the strongest predictor of biochemical failure (Table 1). In a multivariate Cox proportional hazards analysis that considered each factor as a continuous variable, the strongest predictor of failure was the PAP level (P = 0.0001), followed by the Gleason score (P = 0.13), and the PSA level (P = 0.04).

Figure 4.

Freedom from biochemical failure for patients (pts) who were treated with palladium 103 (Pd-103) plus external beam radiation therapy (EBRT) stratified according to prostate specific antigen (PSA) levels (in ng/mL), Gleason score, and prostatic acid phosphatase (PAP) measurement (in Units).

Table 1. Univariate and Multivariate Comparison of Pretreatment Prostate Specific Antigen Levels, Gleason Scores and Prostatic Acid Phosphatase Levels in Predicting Long-Term Biochemical Control
AnalysisPSAGleasonPAP
  1. PSA: prostate specific antigens. PAP: prostatic acid phosphatase.

Univariate0.060.0001.0002
Multivariate0.040.130.0001

PAP measurement was especially helpful in stratifying patients with pretreatment PSA levels between 4 ng/mL and 20 ng/mL, for whom the prognosis is not different when they are subdivided into PSA categories (see Fig. 1). When the PAP subgroup analysis was limited to this relatively favorable group, there was a wide range of prognoses (Fig. 5).

Figure 5.

Biochemical disease control for patients (pts) who had prostate specific antigen (PSA) levels < 20 ng/mL. Note the wide variation in prognosis, despite the fact that the pretreatment PSA level fails to stratify at < 20 ng/mL (see Fig. 3). EBRT: external beam radiation therapy; PAP: prostatic acid phosphatase (in Units); Pd-103: palladium 103.

DISCUSSION

Patients who have prostate carcinoma with pretreatment PSA levels > 10 ng/mL or Gleason scores ≥ 7 generally have not fared as well as patients with lower indices.11 We are encouraged that aggressive, brachytherapy-based treatment seems to provide long-lasting biochemical disease control, even in patients with PSA levels > 20 ng/mL, using the strict biochemical criterion of a posttreatment PSA level ≤ 0.2 ng/mL to consider patients disease free. Although longer follow-up is needed, the apparent plateau on the freedom from failure curves suggests that malignant prostate tumors have been eradicated rather than simply suppressed. Not all brachytherapy series have shown such encouraging results. We attribute our favorable results, in part, to our efforts to achieve wide treatment margins using highly peripheral and extracapsular source placement to achieve a minimum 5-mm treatment margin.11, 13–16

The fact that PAP was the strongest predictor of long-term biochemical failure in patients with otherwise higher risk features reported here suggests that it may be a more accurate indicator of micrometastatic disease compared with the PSA level and the Gleason score. Its potential role, in addition to PSA, recently has been emphasized for patients who undergo radical prostatectomy.3, 4 Similarly, PAP was identified as a significant predictor of disease specific survival in a recent Radiation Therapy Oncology Group analysis.16 This report adds to the rationale for reintroducing PAP measurement into general practice.

The statistically superior discriminatory capability of PAP may be especially important for patients with pretreatment PSA levels between 10 ng/mL and 20 ng/mL. Traditionally, these patients were treated with supplemental beam radiation therapy, with the presumption that their disease was more likely to have an extracapsular component. The results summarized here raise the possibility of using PAP measurement to differentiate patients who have a worse prognosis among patients with PSA levels of 0–20 ng/mL and Gleason scores of 2–7. Patients who have higher pretreatment PAP measurements may be better candidates for other therapies. Considering the greater likelihood of biochemical failure in patients who have elevated pretreatment PAP measurements, as described in previous studies, we have begun adding an additional 540–1080 centigrays (cGy) to the periprostatic tissue, blocking the implant to the 1000 cGy isodose line.

Acknowledgements

The authors thank Ayled Piniero for assistance with data management.

Ancillary