RADIOTHERAPY FOR ISOLATED SERUM PROSTATE SPECIFIC ANTIGEN ELEVATION AFTER PROSTATECTOMY FOR PROSTATE CANCER

0022-5347/00/1633-0845/0 THE JOURNAL OF UROLOGY® Copyright © 2000 by AMERICAN UROLOGICAL ASSOCIATION, INC.®

Vol. 163, 845– 850, March 2000 Printed in U.S.A.

RADIOTHERAPY FOR ISOLATED SERUM PROSTATE SPECIFIC ANTIGEN ELEVATION AFTER PROSTATECTOMY FOR PROSTATE CANCER THOMAS M. PISANSKY, TIMOTHY F. KOZELSKY, ROBERT P. MYERS, DAVID W. HILLMAN, MICHAEL L. BLUTE, STEVEN J. BUSKIRK, JOHN C. CHEVILLE, ROBERT G. FERRIGNI AND STEVEN E. SCHILD From the Division of Radiation Oncology, Departments of Urology, and Laboratory Medicine and Pathology, and Cancer Center Statistical Unit, Mayo Clinic, Rochester, Minnesota, Department of Radiation Oncology, Mayo Clinic, Jacksonville, Florida, and Departments of Urology and Radiation Oncology, Mayo Clinic, Scottsdale, Arizona

ABSTRACT

Purpose: Elevated serum prostate specific antigen (PSA) may be the initial and only indication of disease recurrence after prostatectomy for prostate cancer. External beam radiotherapy may be given in this setting in an attempt to eradicate the disease but therapeutic outcomes after this approach require further description. We describe the intermediate term outcome in a large group of patients treated with radiotherapy and identify pre-therapy factors associated with disease outcome. Materials and Methods: We retrospectively studied a cohort of 166 consecutive patients treated with radiotherapy between July 1987 and May 1996. The Kaplan-Meier method was used to describe patient outcome for the overall study group, and statistical associations of pre-therapy variables with outcome were sought to identify predictive factors. Results: At a median followup of 52 months 46% (95% confidence interval 38 to 55) of patients were expected to be free of biochemical relapse 5 years after radiotherapy. Multivariate analysis identified pathological classification (seminal vesicle invasion), tumor grade and preradiotherapy serum PSA as independent factors associated with biochemical relapse. Although in 1 of 6 patients a chronic complication was attributed to radiotherapy, it was often mild and self-limited in nature. Conclusions: In our current series approximately half of the patients treated with radiotherapy for an isolated elevation of serum PSA after prostatectomy were free of biochemical relapse at 5 years of followup. Radiotherapy may be given in this setting with modest long-term morbidity. KEY WORDS: prostate, prostate-specific antigen, prostatectomy, prostatic neoplasms, radiotherapy

Radical prostatectomy is commonly performed in American men as the primary and only treatment of clinically localized prostate cancer. Postoperative serum prostate specific antigen (PSA) monitoring may provide the initial and only evidence of persistent or recurrent disease in some patients.1 Although to our knowledge the optimal treatment of patients with measurable postoperative serum PSA has not been established, external beam radiation therapy has been given in this setting. Several studies have previously described patient outcomes after this treatment2–12 but many such observational series included a small number of patients who were often observed for a relatively short time. We describe the intermediate term outcome in a large group of patients treated with radiotherapy and identify pre-therapy factors associated with disease outcome in this situation. MATERIALS AND METHODS

Between July 1987 and May 1996, 166 consecutive patients with pathological stage T2 to 3N0,X M013 prostatic adenocarcinoma were treated at our institution with external beam radiation therapy for a detectable postoperative serum PSA of 0.2 ng./ml. or greater. These patients had previously

undergone radical retropubic prostatectomy with (31) or without (133) neurovascular preservation, or transperineal prostatectomy (2) between June 1983 and September 1995 for disease clinically classified as stage T1 in 22, T2 in 118, T3 in 18 and Tx in 8. Preoperative PSA ranged from 2.3 to 219 ng./ml. (median 17.0, mean 23.5). Of the patients 148 (89%) and 18 (11%) underwent surgery at our institution and elsewhere, respectively. All men treated via the retropubic approach underwent pelvic lymph node dissection, and histological examination did not reveal tumor involvement of the recovered lymph nodes. In the 2 remaining patients pathological nodal staging was not done. Seven patients (4%) received transient androgen deprivation therapy with luteinizing hormone releasing hormone agonist and/or an antiandrogen agent after prostatectomy, which was discontinued 0 to 14 months (median 5.6, mean 5.7) before radiation therapy was begun. Table 1 shows additional characteristics of the study group. Disease evaluation before radiotherapy included a history, physical examination, complete blood count, and serum chemistry and PSA determinations as well as chest radiography in 145 patients, radionuclide bone scan in 152 and skeletal radiography in 1. PSA ranged from 0.2 to 15.3 ng./ml. (median 0.9, mean 1.3). Computerized tomography (CT) of the regional lymph nodes in 72 patients (43%) showed no involvement (N0). Nodal status was considered unknown (NX) in the remaining patients since CT performed only for

Accepted for publication September 3, 1999. Editor’s Note: This article is the fifth of 5 published in this issue for which category 1 CME credits can be earned. Instructions for obtaining credits are given with the questions on pages 968 and 969. 845

846

RADIOTHERAPY FOR ISOLATED SERUM PROSTATE SPECIFIC ANTIGEN ELEVATION TABLE 1. Patient characteristics No. Pts. (%)

Overall Pathological tumor classification: T2 T2⫹ (pos. margin) T3a T3b Tx Tumor grade: 2 3–4 Unknown Gleason score: 3–4 5–6 7 8–9 Unknown Preop. PSA (ng./ml.): 4.0 or Less 4.1–10.0 10.1–20.0 20.1–50.0 Greater than 50 Unknown Pre-radiotherapy PSA (ng./ml.): 1.0 or Less 1.1–4.0 4.1–10.0 Greater than 10.0 Radiotherapy dose (cGy.): Less than 6,400 6,400 or Greater DNA ploidy: Diploid Aneuploid Tetraploid Unknown

166 (100) 35 22 56 51 2

(21) (13) (34) (31) (1)

57 (34) 106 (64) 3 (2) 9 41 50 26 40

(5) (25) (30) (16) (24)

4 40 41 52 11 18

(2) (24) (25) (31) (7) (11)

96 (58) 64 (39) 4 (2) 2 (1) 67 (40) 99 (60) 68 (41) 15 (9) 48 (29) 35 (21)

radiotherapy treatment planning did not include the whole contents of the true pelvis. Radiation therapy was started 0.2 to 85 months (median 13.0, average 17.5) after prostatectomy at patient age 44 to 79 years (median 68). It was administered with 6, 10 or 18 MV. photons through 4 to 8 stationary conformal and rotational fields in 159 and 7 patients, respectively. Prostatic fossa target volume was designed with CT based treatment planning and/or on the basis of surgical clip placement. Median dose was 6,400 cGy. (mean 6,300, range 5,400 to 7,240) in 180 to 200 cGy. daily fractions specified at the isocenter. Four patients received 4,500 to 5,040 cGy. at 180 cGy. per fraction to the pelvic lymphatics with 10 or 18 MV. photons via a 4 field approach. Patients were typically evaluated every 3 to 4 months after radiotherapy for 2 years and at 6 to 12-month intervals thereafter. Each evaluation included a history, physical examination, and serum PSA and chemistry determinations. Additional diagnostic studies were performed as indicated by examination findings or assessment of post-therapy serum PSA. Sites of initial tumor relapse were classified as local (prostatic fossa), regional nodal13 and distant (all other sites). Local regional and distant tumor relapse that occurred simultaneously, that is within a 3-month interval, was classified as distant. After radiation therapy PSA was measured to define biochemical relapse. Biochemical relapse was diagnosed when 2 or more post-radiotherapy PSA values were 0.3 ng./ml. or greater and exceeded the post-radiotherapy PSA nadir, when initial post-radiotherapy PSA was greater than that immediately before radiation therapy or when androgen deprivation therapy was iniatiated for a single elevated postradiotherapy PSA value. The date of biochemical relapse was defined as the date of the initial post-radiotherapy PSA elevation or the date androgen deprivation therapy was begun. Tumor grade was determined from the prostatectomy specimen, and was assigned in accordance with the Mayo grading

system14 and/or by the degree of glandular differentiation in association with the pattern of stromal tumor growth, as described by Gleason. For patients who underwent surgery at our institution the radical prostatectomy specimen was promptly submitted and measured. The external surface was inked and then the whole specimen was placed into neutral buffered formalin for overnight fixation. The apical and bladder base portions were initially removed, and the remainder of the prostate was sectioned perpendicular to the longitudinal (apical-basal) axis. The seminal vesicles were sectioned in a similar manner parallel to their junction with the prostate. For patients who underwent surgery elsewhere the specimen was submitted to our institution. Histological examination was performed with special attention to capsular invasion, capsular perforation (tumor extension into periprostatic adipose tissue), surgical margin status and seminal vesicle invasion. Pelvic lymphadenectomy tissue was submitted separate from the prostatectomy specimen and handled as frozen sections in the pathology laboratory. Fibroadipose tissue was dissected, and all lymph nodes were removed, sectioned and histologically examined for the presence and extent of tumor involvement. The number of lymph nodes involved as well as the number recovered was reported with respect to laterality. For the analysis of outcome we used the Kaplan-Meier method to estimate the probability of freedom from disease relapse or survival in the entire study population and in select subgroups.15 Univariate comparison of factors was done with the log rank test. In addition to preoperative and pre-radiotherapy serum (loge) PSA categories, other candidate predictors included pathological tumor classification, tumor grade (2 versus 3 to 4), Gleason score (3 to 4 versus 5 to 6 versus 7 versus 8 to 9), tumor DNA ploidy (diploid versus nondiploid), pre-radiotherapy androgen deprivation therapy and prostatic fossa target volume radiation therapy dose (less than 6,400 cGy. versus 6,400 or greater). Multivariate analysis was done using the Cox proportional hazards model to find the most significant factors associated with disease outcome with variables retained based on a forward procedure and verified with a backward elimination process.16 Because pre-therapy PSA levels were distributed in a nonlinear fashion, individual values were transformed to the natural logarithm (loge PSA) and evaluated as a continuous variable by this model. These methods were used to identify factors that had an independent association with the specified end point. Statistical significance was considered at p ⬍0.05. Time intervals were measured from radiation therapy initiation date to the date of the event under consideration. Time to biochemical relapse was censored at the date of the last PSA determination and disease relapse was censored at the time of the last patient contact or death. RESULTS

After completion of radiation therapy serum PSA decreased in 136 patients (82%). At a median followup of 52 months (mean 56, range 5 to 131) 7 men died of prostate cancer and 4 died of unrelated causes, while 0, 1 and 22 had clinical evidence of local, nodal and distant disease relapse, respectively. The 5-year estimate of overall and clinical relapse-free survival was 93% (95% confidence interval [CI] 89 to 98) and 84% (95% CI 77 to 91), respectively (fig. 1). A total of 89 patients (54%) had biochemical relapse, including 23 with and 66 without clinical disease relapse, while 4 (8%) underwent androgen deprivation therapy for a single elevated post-radiotherapy PSA. In the entire study population the probability of remaining free of biochemical relapse was 46% (95% CI 38 to 55) at 5 years of followup (fig. 1). Univariate log rank analysis identified pathological tumor classification, tumor grade, preoperative serum PSA category and tumor DNA ploidy as factors significantly associated

RADIOTHERAPY FOR ISOLATED SERUM PROSTATE SPECIFIC ANTIGEN ELEVATION

847

FIG. 1. Disease outcome in 166 patients treated with radiotherapy for elevated postoperative serum PSA level. Vertical bars represent 95% CI. OS, overall survival. RFS, clinical relapse-free survival. BCR, biochemical relapse.

TABLE 2. Univariate analysis of candidate predictors of disease relapse at 5 years % Pts. % Pts. Clinical p p Biochemical Relapse-Free Value Value Relapse-Free Survival Pathological tumor classification: T2 T2⫹ (margin positive) T3a T3b Tumor grade: 2 3–4 Gleason score: 3–4 5–6 7 8–9 Preop. PSA (ng./ml.) 0.1–10.0 10.1–20.0 20.1–50.0 Greater than 50 Pre-radiotherapy PSA (ng./ml.) 0.1–1.0 1.1–4.0 Greater than 4.0 DNA ploidy: Diploid Nondiploid Radiotherapy dose (cGy.): Less than 6,400 6,400 or Greater

54 63 51 26

0.05

97 95 88 66

0.01

63 37

0.001

98 75

0.002

0.07

100 85 78 77

0.40

0.04

86 82 78 100

0.49

53 34 50

0.08

91 74 83

0.06

53 39

0.02

92 71

0.002

36 56

0.18

75 92

0.03

76 51 50 28 69 46 32 40

with biochemical relapse (table 2). Pathological tumor classification, tumor grade, tumor DNA ploidy and radiation therapy dose category were associated with clinical relapsefree survival (table 2). Analysis considering the association of factors with overall patient survival was not done due to limitations inherent in the small number of events (deaths). When evaluated as a categorical variable with the log rank test, pre-radiotherapy serum PSA was marginally associated with freedom from biochemical relapse and clinical relapsefree survival (p ⫽ 0.08 and 0.06, respectively, table 2). However, a stronger association was observed with biochemical relapse (p ⫽ 0.0008) and clinical relapse-free survival (p ⫽ 0.007) when PSA was analyzed as a continuous variable by the Cox regression model. Because some factors may exhibit interdependence, multi-

variate analysis was performed to determine which variable(s) retained significance. This analysis demonstrated that pathological tumor classification (T3b),13 tumor grade (3 to 4)14 and pre-radiotherapy serum (loge) PSA (continuous) were important pre-therapy factors independently predictive of biochemical relapse-free status after radiation therapy. Table 3 lists hazards ratios for biochemical relapse and tests of statistical significance for these factors. Kaplan-Meier estimates of remaining free of biochemical relapse according to pathological tumor classification (fig. 2),13 tumor grade (fig. 3)14 and pre-radiotherapy PSA (fig. 4) also illustrate these findings. Since the Gleason grading system is commonly used in contemporary practice but was not available for all study subjects, a Cox regression multivariate analysis that did not include Mayo tumor grade as a candidate predictor was performed using the 126 patients with available Gleason score information. In this scenario Gleason score category (2 to 4 versus 5 to 7 versus 8 to 9) was independently predictive of biochemical relapse-free status (p ⫽ 0.02). Chronic complications of radiation therapy were defined as those persisting beyond or developing more than 90 days after therapy was started. These complications were retrospectively graded according to the late radiation morbidity scoring scheme recommended by the Radiation Therapy Oncology Group and the European Organization for Research and Treatment of Cancer. We identified 38 radiotherapy induced chronic complications in 28 patients (17%). The most common complication was associated with rectal symptoms in 15 patients (9%). In 9 cases it involved mild and in 4 moderate tenesmus, rectal mucous discharge or grade 1 to 2 bleeding that usually did not require intervention. However, hematochezia was treated with laser fulguration in 1 case

TABLE 3. Multivariate regression analysis of pre-therapy factors associated with biochemical relapse-free status Hazards Ratio*

p Value

Pathological tumor classification: T3a 0.95 0.85 T3b 1.70 0.05 Tumor grade 3–4 2.16 0.002 Pre-radiotherapy PSA 1.43 0.003 * Likelihood ratio test p value from the Cox regression model expressed relative to baseline (hazards ratio 1.0) according to pathological tumor stage T2, tumor grade 2 or pre-radiotherapy PSA value 1 unit smaller on the loge scale.

848

RADIOTHERAPY FOR ISOLATED SERUM PROSTATE SPECIFIC ANTIGEN ELEVATION

FIG. 2. Biochemical relapse-free probability after radiotherapy according to pathological tumor classification.13 Vertical bars represent 95% CI.

FIG. 3. Biochemical relapse-free probability after radiotherapy according to tumor grade.14 Vertical bars represent 95% CI

and bowel diversion was required in another. In addition, 12 patients (7%) had small and/or large intestinal symptoms, including mild and moderate diarrhea that required dietary modification in 11 and 1, respectively. Four patients (2%) had cystitis that was associated with gross self-limiting hematuria in 2. Urethral stricture in 6 cases (4%) was managed by dilation in 4. Urinary incontinence in 1 case was attributed to radiation therapy. DISCUSSION

Persistent or recurrent detectable serum PSA may be observed after radical prostatectomy for prostate cancer,1 which in the absence of clinically evident disease may be referred to as biochemical relapse. Although biochemical relapse may prompt further diagnostic investigation, evaluations, such as physical examination, transrectal ultrasonography, bone scintigraphy and CT, have limited value for identifying disease site(s). Irregularities of the prostatic fossa may be apparent on digital rectal examination17 but this finding does not correlate strongly with biopsy results.17, 18

Scar tissue and tumor recurrence may appear hypoechoic on transrectal ultrasonography,19 which may explain the low diagnostic specificity of this procedure.19 In our study patients were evaluated in a manner consistent with contemporary practice guidelines,20 and clinical examination with or without diagnostic imaging tests did not demonstrate evidence of metastatic disease. The outcome in cases selected for radiotherapy for postoperative biochemical relapse has been exclusively described in observational case reports2–12, 21 because to our knowledge prospective comparative trial results do not exist in this setting. Compared with previous studies our investigation involved a consecutive and relatively large patient cohort with extended followup. Although 4% of patients received temporary androgen deprivation therapy postoperatively, it was discontinued before radiation therapy and did not affect disease outcome (data not shown). Of the patients 98% were treated in a relatively homogeneous manner with prostatic fossa radiation therapy as commonly recommended,20 which

RADIOTHERAPY FOR ISOLATED SERUM PROSTATE SPECIFIC ANTIGEN ELEVATION

849

FIG. 4. Biochemical relapse-free probability after radiotherapy according to serum PSA before radiotherapy. Vertical bars represent 95% CI.

allowed a straightforward description of radiotherapeutic outcome for this condition. After completing radiotherapy serum PSA decreased in 82% of our patients, which is comparable to the approximate 75% (range 51% to 93%) response rate in previously published studies (table 4).2–7, 9, 12, 21 These findings indicate that locally persistent disease was at least partially responsible for the serum PSA level in most patients. However, a substantial proportion of patients with an initially favorable biochemical response to radiation therapy subsequently had an increasing serum PSA profile. Of the patients 46% were expected to remain biochemical relapse-free 5 years after radiotherapy. However, this outcome favorably compared with that of previous reports with similar followup (table 4).2–12, 21 In our series the predominant clinical pattern of disease relapse resulted from hematogenous dissemination (96% of all relapses), which was evident in 13% of patients during post-therapy followup. Although this observation suggested that occult systemic disease was present in some patients, unrecognized regional nodal involvement,22 or tumor persistence in or near the irradiated area must also be considered a source of biochemical relapse. After the study period 111indium labeled capromab pendetide radioscintigraphy was introduced for use in patients with biochemical relapse after prostatectomy for prostate cancer. Although radioscintigraphy findings may be associated with radiotherapeutic outcome, patients with uptake outside the prostatic fossa may also have complete biochem-

ical response to locally administered radiation therapy.23 In addition, the 49% sensitivity and 71% specificity of this test,24 which was defined in patient groups with a broad range of postoperative PSA, may not be sufficient to influence medical decision making. Further study of this imaging approach for patients with lower serum PSA elevations would appear prudent before this approach is broadly applied to those who may otherwise be considered candidates for radiation therapy. In our study pathological stage (seminal vesicle invasion), tumor grade and pre-radiotherapy serum PSA were identified by multivariate analysis as independent predictors of biochemical relapse after radiation therapy. Although a consistent association of pathological stage with outcome was not evident in previous studies, its importance has been suggested by others9, 10, 12 and it clearly correlated with seminal vesicle invasion in our study. Likewise to our knowledge the prognostic significance of tumor grade in this setting has not been previously established but it was evident in our series. Pre-radiotherapy serum PSA has consistently been recognized as a factor associated with post-therapy biochemical relapse,4, 7–10, 12, 21 and its importance was confirmed by our investigation. However, unlike pathological stage and tumor grade, the risk of relapse increased in a continuous, and not stepwise manner with serum PSA, which created difficulty in selecting an optimal cutoff point for this factor. Although previous groups used serum PSA cutoff values of 1.0,7, 12 1.6,10 2.09 and 2.54 ng./ml., to our knowledge the

TABLE 4. Radiation therapy for elevated PSA after prostatectomy References

No. Pts.

% Observed Biochemical Response

82 — Cadeddu et al*, 11 45 Approximately 67 Coetzee et al*, 5 *, 8 48 — Crane et al 60 92 Do et al12 *, 21 103 Greater than 51 Eulau et al 47 Greater than 60 Forman et al9 3 17 65 Haab et al 37 Greater than 60 McCarthy et al2 6 37 70 Medini et al 48 — Morris et al*, 10 7 46 93 Schild et al 4 53 59 Wu et al Present series 166 82 * Includes patients treated adjuvantly or for palpable local tumor recurrence.

% Biochemical Relapse-Free 10 51 24 55 34 64 18 Approximately 30–67 27 68 50 26 46

End Point 5-Yr. actuarial Observed rate 5-Yr. actuarial 5-Yr. actuarial 5-Yr. actuarial Observed rate Observed rate 3-Yr. actuarial Observed rate 3-Yr. actuarial 3-Yr. actuarial 2-Yr. actuarial 5-Yr. actuarial

850

RADIOTHERAPY FOR ISOLATED SERUM PROSTATE SPECIFIC ANTIGEN ELEVATION

usefulness of such values in medical decision making has not been determined. To date it would appear that when radiation therapy is given for postoperative biochemical relapse, it should be instituted at the earliest possible time rather than assuming that a similar outcome will result in patients with PSA below a certain level (for example 1 ng./ml.). Although it was not uniformly observed,4, 7, 8, 10 several reports suggest that persistent serum PSA elevation after radical prostatectomy portends a worse biochemical relapsefree outcome after radiation therapy than when PSA is initially undetectable after prostatectomy but subsequently increases.9, 21 Because serum PSA was not determined in all patients in our study, we did not assess when a nondetectable nadir may have occurred (for example after 5 serum halflives). For similar reasons analysis was not extended to evaluate the possible prognostic importance of anastomotic biopsy status (positive versus negative),6 which may reveal a false-negative result,18 or PSA doubling time after radical prostatectomy.4 Radiotherapy is generally associated with a low rate of chronic morbidity. Although 1 in 6 patients in our cohort had a late complication, it was generally mild and responsive to conservative measures. Other studies have also indicated that urinary incontinence in 0% to 6% of cases,4, 6, 7, 9, 10, 12 mild to moderate proctopathy in 0% to 17%4, 6, 7, 9, 10, 12 and chronic cystitis in 0% to 13%4, 6, 7, 10 may result from radiotherapy in this setting. Nevertheless, the reported complication rate is relatively low and of a modest degree. Although single institution observational case series of the type that we describe may provide useful information, controlled clinical trials are necessary to define the optimal management approach in this clinical setting. The European Organization for Research and Treatment of Cancer study 30943 is investigating the role of immediate versus delayed androgen deprivation in asymptomatic patients with biochemical relapse after prostatectomy and/or radiotherapy, and the Radiation Therapy Oncology Group study 9601 is testing the role of the nonsteroidal antiandrogen bicalutamide in association with prostatic fossa radiotherapy. The successful completion of prospectively randomized trials of these types is required to improve medical community understanding of this condition and provide further direction for the care of these patients. CONCLUSIONS

In our series approximately half of the patients treated with radiotherapy for an isolated serum PSA elevation after prostatectomy were free of biochemical relapse at 5 years of followup. Radiotherapy in this setting may be given with modest long-term morbidity. REFERENCES

1. Murphy, G. P., Mettlin, C., Menck, H. et al: National patterns of prostate cancer treatment by radical prostatectomy: results of a survey by the American College of Surgeons Committee on Cancer. J Urol, part 2, 152: 1817, 1994 2. McCarthy, J. F., Catalona, W. J. and Hudson, M. A.: Effect of radiation therapy on detectable serum prostate specific antigen levels following radical prostatectomy: early versus delayed treatment. J Urol, 151: 1575, 1994 3. Haab, F., Meulemans, A., Boccon-Gibod, L. et al: Effect of radiation therapy after radical prostatectomy on serum prostatespecific antigen measured by an ultrasensitive assay. Urology, 45: 1022, 1995

4. Wu, J. J., King, S. C., Montana, G. S. et al: The efficacy of post-prostatectomy radiotherapy in patients with an isolated elevation of serum prostate-specific antigen. Int J Radiat Oncol Biol Phys, 32: 317, 1995 5. Coetzee, L. J., Hars, V. and Paulson, D. F.: Postoperative prostate-specific antigen as a prognostic indicator in patients with margin-positive prostate cancer, undergoing adjuvant radiotherapy after radical prostatectomy. Urology, 47: 232, 1996 6. Medini, E., Medini, I., Reddy, P. K. et al: Delayed/salvage radiation therapy in patients with elevated prostate specific antigen levels after radical prostatectomy. A long term follow-up. Cancer, 78: 1254, 1996 7. Schild, S. E., Buskirk, S. J., Wong, W. W. et al: The use of radiotherapy for patients with isolated elevation of serum prostate specific antigen following radical prostatectomy. J Urol, 156: 1725, 1996 8. Crane, C. H., Rich, T. A., Read, P. W. et al: Preirradiation PSA predicts biochemical disease-free survival in patients treated with postprostatectomy external beam irradiation. Int J Radiat Oncol Biol Phys, 39: 681, 1997 9. Forman, J. D., Meetze, K., Pontes, E. et al: Therapeutic irradiation for patients with an elevated post-prostatectomy prostate specific antigen level. J Urol, 158: 1436, 1997 10. Morris, M. M., Dallow, K. C., Zietman, A. L. et al: Adjuvant and salvage irradiation following radical prostatectomy for prostate cancer. Int J Radiat Oncol Biol Phys, 38: 731, 1997 11. Cadeddu, J. A., Partin, A. W., DeWeese, T. L. et al: Long-term results of radiation therapy for prostate cancer recurrence following radical prostatectomy. J Urol, 159: 173, 1998 12. Do, T., Parker, R. G., Do, C. et al: Salvage radiotherapy for biochemical and clinical failures following radical prostatectomy. Cancer J Sci Am, 4: 324, 1998 13. Prostate. In.: AJCC Cancer Staging Manual, 5th ed. Edited by I. D. Fleming, J. S. Cooper, D. E. Henson et al. Philadelphia: Lippincott-Raven, p. 219, 1997 14. Utz, D. C. and Farrow, G. M.: Pathologic differentiation and prognosis of prostatic carcinoma. JAMA, 209: 1701, 1969 15. Kaplan, E. L. and Meier, P.: Nonparametric estimation from incomplete observations. J Am Stat Assoc, 53: 457, 1958 16. Cox, D. R.: Regression models and life tables. J Roy Stat Soc, 34: 187, 1972 17. Lightner, D. J., Lange, P. H., Reddy, P. K. et al: Prostate specific antigen and local recurrence after radical prostatectomy. J Urol, 144: 921, 1990 18. Fowler, J. E., Brooks, J., Pandey, P. et al: Variable histology of anastomotic biopsies with detectable prostate specific antigen after radical prostatectomy. J Urol, 153: 1011, 1995 19. Connolly, J. A., Shinohara, K., Presti, J. C., Jr. et al: Local recurrence after radical prostatectomy: characteristics in size, location, and relationship to prostate-specific antigen and surgical margins. Urology, 47: 225, 1996 20. Decision trees and management guidelines. Patterns of Care Committees. Semin Radiat Oncol, 7: 163, 1997 21. Eulau, S. M., Tate, D. J., Stamey, T. A. et al: Effect of combined transient androgen deprivation and irradiation following radical prostatectomy for prostatic cancer. Int J Radiat Oncol Biol Phys, 41: 735, 1998 22. Ferrari, A. C., Stone, N. N., Eyler, J. N. et al: Prospective analysis of prostate-specific markers in pelvic lymph nodes of patients with high-risk prostate cancer. J Natl Cancer Inst, 89: 1498, 1997 23. Kahn, D., Williams, R. D., Haseman, M. K. et al: Radioimmunoscintigraphy with In-111-labeled capromab pendetide predicts prostate cancer response to salvage radiotherapy after failed radical prostatectomy. J Clin Oncol, 16: 284, 1998 24. Kahn, D., Williams, R. D., Manyak, M. J. et al: 111Indiumcapromab pendetide in the evaluation of patients with residual or recurrent prostate cancer after radical prostatectomy. J Urol, 159: 2041, 1998