Outcome for Surgically Staged Localized Prostate Cancer Treated With External Beam Radiation Therapy

o(m-5347~7/lS75-17~3.W0

V O ~157, . 1754-1759, May 1997 Printed in V . S A

THEJOUR~AL OF UROLCCY Copyright 0 1997 by -CAN UROLCCICAL AsSOcumON, h c .

OUTCOME FOR SURGICALLY STAGED LOCALIZED PROSTATE CANCER TREATED WITH EXTERNAL BEAM RADIATION THERAPY CURT R. POWELL,* THOMAS K. HUISMAN, ROBERT H. RIFFENBURGH, ERIC L. SAWDERS, KELLY J. BETHEL

AND

PETER A. S. JOHNSTONE

From the Departments of Urology, Laboratory (Pathology Division), Clinical Investigation and Radiology (Radiation O ~ o l o g Division), y Naual Medical Center, and Radiation Oncology Division, University of California, San Diego, California

ABSTRACT

Purpose: A retrospective analysis was performed on patients with surgically staged localized prostate cancer treated with external beam radiation therapy for 10-year overall, cause specific and disease-free survivals based on lack of clinical recurrence and 2 separate prostate specific antigen criteria for cure. Materials and Methods: We analyzed 145 patients who received external beam radiation therapy after a negative staging pelvic lymphadenectomy for pxostate cancer. Followup data were available for 129 patients (90%).Disease was stage A in 29 patients (22.5%),stage B in 64 (49.6%),stage B2/C in 2 (17%) and stage C in 14 (10.9%).Average potential followup from date of diagnosis was 11.5 years (minimum 7.2). Of the patients 87 potentially can be followed for longer than 10 years. Disease-free survival was based on a normal digital rectal examination, lack of symptoms suspicious for metastasis and application of 2 separate prostate specific antigen criteria of 4 ng./ml. or less (group l),or 1.5 ng./ml. or less (group 2). Survival was analyzed with the Kaplan-Meier actuarial method. Results: Actuarial overall survival at 10 and 15 years was 63.7 and 49.6, respectively, and cause specific survival was 84.2 and 80%, respectively. Disease-free survival was 54.5 and 32.4%, respectively, for group 1, and 42.3 and 9.6%,respectively, for group 2. Conclusions: The improved patient selection inherent in surgical staging before definitive external beam radiation therapy provides for improved overall and cause specific survival over that of patients without surgical staging. Biochemical disease-free survival also appears to be improved. KEY WORDS: prostatic neoplasms,

neoplasm staging, radiotherapy

clinical stage T2 (B) prostate cancer it was concluded that there was no clear-cut evidence of superiority for radical prostatectomy, radiation therapy (external beam or interstitial) or surveillance. We analyzed retrospectively 129 patients who received definitive external beam radiation therapy for localized prostate cancer after a negative pelvic lymph node dissection for surgical staging. Outcome was assessed by examining overall and cause specific survivals, as well as biochemical diseasefree survival based on a normal digital rectal examination, lack of symptoms suspicious for metastasis, and 2 different prostate specific antigen (PSA) criteria of 4.0 less than or equal to mg./ml. and 1.5 ng./ml. Our institutional policy from the mid 1970s to the mid 1980s called for surgical staging before definitive radiotherapy for clinically localized prostate cancer. Between November 1974 and August 1988,164 men with clinical stages A to C prostate cancer underwent staging pelvic lymphadenectomy as described by Herr.5 No patient died or had significant morbidity after this procedure. Of the patients 19 (12.8%) had positive lymph nodes, leaving 145 with surgically staged localized prostate cancer treated with radiotherapy. Information on these patients was obtained from inpatient and outpatient charts, the tumor registry, patient questionnaires and telephone contact. Patients who were alive and in the San Diego area were followed at our clinic. 5000. We contacted the urologists caring for patients living elseEditor-%Note: "his article is the fourth of 5 published in this iesue for which category 1 CME credits can be earned. In- where to obtain information about the disease status, includstructions for obtaining credits are given with the questions ing digital rectal examination, PSA and other relevant tests, on pages 1910 and 1911. such as bone scans. Data are reported as of August 1995 and

Comparisons among series reporting outcome for treatment of localized prostate cancer with surgery or radiation therapy are difficult for several reasons. Results from previous radiation oncology series are potentially skewed in that the status of the pelvic lymph nodes was not known at treatment in the majority of patients. It has been reported that 10 to 20%of patients with stage T1, 30% with stage T2 and 40 to 60% with stage T3-4 disease will have positive pelvic lymph nodes.'.* Potential modifiers of results include the well documented variability between surgical and clinical staging, and under grading of biopsy specimens.3 The recent findings of the Prostate Cancer Clinical Guidelines Panel reported by Middleton et al showed that differences among treatment series regarding variables, such as age, tumor grade, pelvic lymph node status, number of patients and length of followup, were too great to allow valid comparisons.4 ARer reviewing outcomes data for therapy of Acce ted for publication November 1, 1996. Reaci'at annual meeting of American Urological Association, Orlando, Florida, May 4-9, 1996. The Chief, Bureau of Medicine and Surgery, Navy Department, Washington, D. C., Clinical Investigation Program sponsored this report #S-94-126,as required by HSETCINST 6000.41A.The opinions and assertions contained herein are those of the authors and are not to be construed as official or re resenting the views of the United States Navy or Department of Defense. * Re uests for reprints: Clinical Investigation De artment, 34800 Bob Wlson Dr., Naval Medical Center, San Diego, 8alifornia 92134-

1754

PROSTATE CANCER PATIENTS TREATED WITH EXTERNAL BEAM RADIATION THERAPY

followup is continuing. Of these 145 patients 16 (11%) were lost to followup leaving 129 who comprised our study group. m e disease was stage A in 29 patients (22.5%),stage B in 64 (49.6%),stage B2/C in 22 (17%)and stage C in 14 (10.9%).Of the patients lost to followup the disease was stage A in 3 (18.8%,),stage B in 9 (56.3%),stage BWC in 1(6.3%)and stage C in 3 (18.8%). Average patient age at diagnosis and treatment was 66.7 years (range 50 to 81) and average potential followup from date of diagnosis was 11.5 years (median 10.9, range 7.2 to 20.9).Tumors were graded pathologically using the Byar and Mostofi grading system.6 When possible, the original pathological blocks were obtained and reviewed by 1 pathologist (K. J. B.), who was blinded to outcome and who assigned Gleason scores? There were 72 specimens (56%)available for review. A Gleason score of 2 to 4 indicated well (36 cases), 5 to 6 moderately (48) and 8 to 10 poorly (22) differentiated disease. Differentiation was unknown in 2 cases. Gleason scores of 7 (21 cases) were analyzed separately. PSA data were obtained for 102 patients (79.1%)from as early as 1987, although PSA did not become widely used at our institution until 1989 (table 1).Values were generated with the Hybritech* antibody sandwich-type assay before 1993 when the PSA protocol was changed to a micro-particle enzyme immunoassay. Overall, 33 patients (25.6%) were treated with hormones, salvage prostatectomy and/or cryoablation after treatment for clinical disease progression as evidenced by an increasing PSA or positive bone scan. Of these patients 5 had recurrent disease on a channeling transurethral prostatectomy specimen (1)or positive transrectal ultrasound guided biopsy (4). The average time to positive biopsy was 90.4 months (range 14 to 239). Radiotherapy was delivered with a median dose of 66.7 Gy. (range 63 to 70.2) within a median of 55 days (range 34 to 95), generally treating only the prostatic bed with opposed anteroposterior, 4-field or arc techniques. Patients were considered to have clinically recurrent disease after radiotherapy if they had a new abnormality on digital rectal examination, positive transrectal ultrasoundguided biopsy, positive bone scan or signs of local progression, such as ureterovesical obstruction or positive channeling transurethral resection of the prostate biopsy. Outcome for the entire group was evaluated for overall survivalpatients dying of any cause were regarded as having treatment failure at death and cause specific survival-patients dying of prostate cancer were regarded as having treatment failure at death, while those dying of intercurrent causes were censored a t death. For patients with PSA data available biochemical disease-free survival was determined in those who had never received hormonal therapy, no clinical evidence of recurrence and a PSA of 4.0 (group l) or 1.5 (group 2) ng./ml. or less. Patients who received hormonal therapy, salvage prostatectomy or cryoablation before PSA data were available were included in this analysis, since it is assumed that they would have had PSA failure a t clinical failure. Patients were regarded as having treatment failure when the PSA became greater than or equal to 4.0 ng./ml. in group 1 (those dying of intercurrent causes with a normal PSA were

* Hybritech, San Diego, California.

1755

TABLE1. Latest PSA data in patients according to stage at diagnosis and hormonal therapy Stage

No. M A (ng./ml.)

Hormonal Therapy

Overall

No Yes

A

No

B

Yes No Yes

B2C C

_

-More

_

~

Than 4.0 5 0 8

11

3 22 8 8

Total No. Pts. 18 3

2 0

7 4

37

3

4

1 1

3 2

No

4

2

0

12 7 6

Yes

1 45

1

No

16

2 12

73

YDC

16

5

8

29

15

4

censored at death) and greater than or equal to 1.5 ng./ml. in group 2 (those dying of intercurrent causes with a PSA of less than 1.5 ng./ml. were censored at death). For both definitions PSA values exceeding respective thresholds were verified with a second PSA level when possible. It should be noted that calculations for biochemical disease-free survival were based strictly on clinical examination and PSA, and 5 patients were included in the calculation for group 2 whose latest PSA was the first value greater than 1.5 but less than 4.0 ng./ml. The Kaplan-Meier product limit method was used to estimate survival probabilities.8 Comparison among survival curves was performed with the log-rank and paired comparisons tests. RESULTS

Table 2 shows patient status by clinical stage. Overall survival at 10 and 15 years was 63.6 and 49.6%, respectively (fig. 1).There was no significant difference when this finding was stratified by pretreatment clinical stage (p = 0.51) or tumor grade (p = 0.97). Two separate curves were calculated for cause specific survival (fig. 2). In 1group all 13 patients dead of unknown cause were removed from analysis, and the 10 and 15-year survivals were 84.2 and 80%, respectively. In the other group these patients were considered to be dead of prostate cancer, and the 10 and 15-year survivals were 75.6 and 71.8%, respectively. Stratification by stage in the former group showed slight statistical significance (p = 0.041), which was mostly due to the difference between stages A and B tumors. There was no significant difference based on tumor grade (p = 0.451, and in the latter group there was also no significant difference based on stage (p = 0.33) or grade (p = 0.84). Of 93 patients who would now be considered candidates for radical prostatectomy based on clinical stage alone (A and B) 36 (38.7%) are dead of all causes, including 16 (17.2%) dead of prostate cancer and 8 (8.6%) dead of unknown causes. Of the 57 patients in this group still alive 35 (61.4%)are without clinical or biochemical evidence of disease. Of 36 patients who had stages BYC and C cancer 24 (66.7%) are still alive, including 7 (29.2%) without evidence of clinical or biochemical disease. Biochemical disease-free survival for patients with PSA data available is shown in figure 3, A. For all stages survival

TABLE2 . Current patient status No. F’ta. Alive

No. F’ta. Dead Stage A B B2C C

Prostate Ca Intercurrent Disease 1 15 1

2

2 6

1

0

No Evidence of

unknown Causes

nisaaw

Prostate Ca

-._--I-

3

2 1

3 5 3 2

6 16

N~ ~

v i of Disease

12 21

12

4

5 -

3 -

Total d ~ NO.~F’ta.~ ._ B

64

22

14 -

1766

PROSTATE CANCER PATIENTS TREATED WITH EXTERNAL BEAM RADIATION THERAPY 1.00-

Sunrhral Probability 0.75

-

0.60

-

0.25

-

0-

I 0

I 10

I 5

I 15

I 20

Years Since Diagnosis

FIG. 1. Overall survival for all stages 1.00

-

Survival Probability

I

0.76

-

0.60

-

0.25

-

1

I

I

I

I

0

5

10

16

20

Years Since Diagnosis FIG.2. Cause s

c survival. A, 13 patients dead of unknown cause removed from analysis. B , patients dead of unknown cause Considered dead o prostate cancer (log-rank p = 0.12).

p""

at 10 and 15 years was 54.5 and 32.4%, respectively,in group 1, and 42.3 and 9.6%, respectively, in group 2. There was no statistically significant difference when comparing patients by stage for group 1(p = 0.179) or 2 (p = 0.285). Comparisons by grade were sigdicant for both groups (fig. 3, B and C).An examination of the chi-square statistics for individual pairs of groups revealed that for groups 1 and 2 the significance was entirely between the well differentiated tumors and the other groups, with the greatest differencebeing between the well and poorly differentiated lesions. Moderately differentiated, Gleason wore 7 and poorly differentiated tumors were not significantly different from each other.

DISCUSSION

To our knowledge we describe the long-term outcome of the largest series to date of patients with surgically staged disease treated with external beam radiation therapy. Our data were consistent with those of other series (table 31.9-18 Others have evaluated outcome in patients with clinically localized disease that was not surgically staged, and our results are consistent with those of Bagshawl0 and Hanks19 et a1 who found superior results in patients with surgically staged cancer. Surgical series for patients with early stage disease (table

PROSTATE CANCER PATIENTS TREATED WITH EXTERNAL BEAM RADIATION THERAPY

1757

<:k *I-

*--

111-

I

A

Y""

0 -

IL

"II

wu +-

0-

I

WI

B

C

VWShUo*Snac

I 1

"

II

10

m

V-yU-

FIG.3. Biochemical disease-free survival. A, for all stages (log-rank p = 0.027). B , by ade using PSA less than 4.0 ng./ml. (log rank p = 0.008). C, by grade using PSA less t h a n 1.5 ng./ml. (log rank p = 0.011).0, well d i g r e n t i a t e d . I, moderately differentiated. 2, poorly differentiated. NEDI, group 1. NEDl.5, group 2. TABLE3. Comparison of patients treated with radiotherapy for surgically staged and nonsurgically staged prostate cancer. and those treated with radical prostatectomy Reference Stage No. Pts. Type of Trial Actuarial Followup (yrs.) % Overall Survival % Cause Specific Survival Bagshaw et a19 A2-B Bagshaw et all0 Tl-TZ Hanks et al" Tl-TZ Lee and Sause" Tlb-T2 T3

52 60 104 20 15

Hahn et all3

Tla-Tlb TZa-TZb Kuban et all4 A2-Bl B2-C Bagshaw et all0 Tl-TZ

152 346 652

Blute et all' Paulson et all6 Middleton et all6 Zincke et all'

A-B Tl-TZ A-B2

315 441 46

T1-2c

Current study

A-C

Surgicalty staged Ca treated with radiotherapy Single institution, retrospective 13 Single institution, retrospective 15 RTOG 7706 10 Single institution, retrospective 15 Nonsurgically staged Ca treated with radiotherapy Single institution, retrospective 15 Single institution. retrospective

75 70 86 75 22

60 46 67 40 15 75.2 42

10

100-71 57-38 83-93 78-50

Single institution, retrospective 15 Ca treated with radiothemjy Single institution, retrospective 15 Single institution, retrospective 10 Single institution, retrospective 10

40

68

71

93

67

83

1,143

Single institution, retrospective

129

Single institution, retrospective

75 60 64 50

90 83 76-84 72-80

690

-

10 15 10 15

88

3) appeared to show slightly improved survival results, alThe fact that there was no significant difference in overall though it has been noted that patients in such series are or cause specific survival based on stage or grade of the generally younger and in better health than those in radia- tumor and that there was no significant difference in biotion therapy series.19Given the lack of pathological staging of chemical disease-free survival based on stage indicates the the prostate, it is not possible to compare adequately our data difficulties in performing a retrospective analysis of patients to these of surgical series. However, based on clinical stage evaluated by more than 1physician and highlights the inacalone, the majority of our patients (72%with stages A and B2 curacies of clinical staging. Grading of the tumors was also disease) would be considered surgical candidates today at our not standardized. Only 56% of the specimens were available institution. for a retrospective review by 1blinded pathologist, and in 40 It is noteworthy that when calculating cause specific sur- of 72 specimens (56%) the Gleason score assigned placed the vival information concerning the cause of death was obtained tumor into a different category than the original evaluation. The level of PSA that defines disease recurrence after from patient charts as well as tumor registry data, and the cause of death was determined by the attending physician a t radiation therapy is unclear. Several investigators have used that time. Since different physicians made this determina- varying PSA criteria to show biochemical disease-free surtion for most patients, there is clearly the potential for some vival that are dramatically less than those for clinical disease-free survival (table 4).14.19-21 Our results for groups significant objective bias in determining cause of death.

Reference

TABLE4. Comparison of current study with radiation series based on biochemical disease-free survival Staee No.Pts. Actuarial Followu~(vrs.) PSA Criteria (ndml.) % Biochemical Disease-Free Survival Less than 4.0 Less than 1.0 Less than 4.0

52 40 35-18 21-11

10

Less than 3.5 Less than 1.5

88

84

10

Less than 4.0

91

10 15 10

Zietman et alZo

Tl-T2

504

10

Kuban et all4

A2-Bl B2-C Tlb-TZ

652

10

17

Hanks et allQ Schellhammer et a121 Current study

A2-B 1 B2-C All stages A 4

Less than 0.5 Less than 4.0 Less than 1.5

65 35-20 20-10 Less than 10 55 32 42

1758

PROSTATE CANCER PATIENTS TREATED WITH EXTERNAL

BEAM RADIATION THERAPY

1 and 2 are potentially skewed for several reaaons. We in- 10. Bagshaw, M.A., Cox, R. S. and Hancock, S . L.:Control of prostate cancer with radiotherapy: long-term results. J. Urol., 152: cluded in the analysis patients who received adjuvant ther1781,1994. apy for clinical recurrence a t a time when PSA was not available. It is possible that the PSA would have been normal 11. Hanks, G. E., Perez, C. A., Kozar, M., Asbell, S. O., Pilepich, M. V. and Pajak, T. F.: PSA confirmation of cure a t 10 years of at that time and including them in the analysis would artiTlb, T2,NO, MO prostate cancer patients treated in RTOG ficially lower the results for disease-free survival. However, protocol 7706 with external beam irradiation. Int. J. Rad. it is unlikely that PSA would have been normal when the paOncol. Biol. Phys., So.289,1994. tients had clinical evidence of disease recurrence. Also, 5 12. Lee, R. J. and Sause, W. T.: Surgically staged patients with patients in group 2 had treatment failure with only 1 PSA prostatic carcinoma treated with definitive radiotherapy: value of greater than 1.5 ng./ml. and no further PSA measfifteen-year results. Urology, 43:640. 1994. urements were available. It is possible that the subsequent 13. Hahn, P.,B a d , E., Cheang, M., Kostrya, J. and Roelss, R.: Long-term outcome of radical radiation therapy for prostatic PSA values would not be increasing, and that including them carcinoma: 1967-1987. Int. J. Rad. Oncol. Biol. Phys., 34: 41, would impact negatively on disease-free survival. However, 1996. including them in the analysis provided for a more conservative estimate of disease-free survival. Finally, our results 14. Kuban, D. A,, el-Mahdi, A. M. and Schellhammer, P. F.: Prostate-specific antigen for pretreatment prediction and postwere potentially inflated by the significant gap between treatment evaluation of outcome after definitive irradiation for treatment and initial measurement of PSA. The majority of prostate cancer. Int. J. Rad. Oncol. Biol. Phys., 3 2 307. 1995. these patients were treated in the pre-PSA era, and a median 15. Paulson, D.F., Moul, J. W. and Walther, P. J.: Radical prostaof 6.4 years had elapsed from treatment to initial PSA meastectomy for clinical stage T1-SNOMO prostatic adenocarciurement. Furthermore, PSA data were unavailable for 27 noma: long-term results. J. Urol., 144: 1180,1990. patients who died. We are further investigating the clinical 16. Middleton, R. G.,Smith, J. A., Jr., Melzer, R. B. and Hamilton, P. E.: Patient survival and local recurrence rate following significance of an elevated PSA in patients with this much radical prostatectomy for prostatic carcinoma. J. Urol.. 136: followup. CONCLUSIONS

To our knowledge this is the largest series to date on the long-term outcome of patients with surgically staged localized prostate cancer treated with definitive radiotherapy. The results were consistent with smaller series showing that surgical staging offers improved survival over clinical staging. Radical prostatectomy series of patients with similar stage cancer showed slightly improved survival but differences in the patient populations among series made accurate comparisons difficult. Although absence of early PSA data may artificially inflate our biochemical disease-free survival results, they would appear to be improved over nonsurgically staged cases, which is consistent with improved patient selection. REFERENCES

1. Banell, W., Bean, M. A., Hilaris, B. S. and Whitmore, W. F., Jr.: Prostatic adenocareinoma: relationship of grade and local extent to the pattern of metastases. J . Urol., 11s: 278,1977. 2. Donohue, R. E., Augspurger, R. R., Mani, J. H., Biber, R. J., Whitesel, J. A., Fauver, H. E., Mohr, S.,Wettlaufer, J. N., Scanavino, D. and pfister, R. R.: Pelvic lymph node dissection. Guide to patient management in clinically, locally confined adenocarcinoma of prostate. Urology, 20: 559,1982. 3. Johnstone, P. A., W e n b u r g h , R., Saunders, E. L. and Willison, F. W.: Grading inaccuracies in diagnostic biopsies revealing prostatic adenocarcinoma: implications for definitive radiation therapy. Int. J. Rad. Oncol. Biol. Phys., 3 2 479, 1995. 4. Middleton, R. G., Thompson, I. M., Austenfeld, M. S., Cooner, W. H., Correa, R. J., Gibbons, R. P., Miller, H. C., Osterling, J. E., Resnick, M. I., Smalley, S. R. and Wasson, J. H.: Prostate cancer clinical guidelines panel summary report on the management of clinically localized prostate cancer. J. Urol., 1&i: 2144,1995. 5. Herr, H. W.: Pelvic lymphadenectomy and iodine-125 implantation. In: Genitourinary Tumors: Fundamental Principles and Surgical Techniques. Edited by D. E. Johnson and M. A. Boileau. New York Grune & Stratton, chapt. 6,pp. 63-73, 1982. 6. Byar, D. P. and Mostofi, F. K: Carcinoma of the prostate: prognostic evaluation of certain pathological features in 208 radical prostatectomies. Examined by the step-section technique. Cancer, 3 0 5, 1972. 7. Gleason, D. F.: Classification of prostatic carcinomas. Cancer Chemother. Rep., M): 125,1966. 8. Kaplan, E. L. and Meier. P.: Nonparametric estimation from incomplete observation. J. Amer. Stat. Ass., 63:457,1958. 9. Bagshaw, M. A., COX,R. S. and Ray, G. R.: Status of radiation treatment of prostate cancer at Stanford University. Natl. Cancer Inst. Monogr., ?: 41,1988.

422, 1986. Zincke, H., Farrow, G. M., Therneau, T. 17. Blute, M.L., Nativ, 0.. and Lieber, M. M.: Pattern of failure after radical retropubic prostatectomy for clinically and pathologically localized adenocarcinoma of the prostate: influence of tumor deoxyribonucleic acid ploidy. J . Urol., 142 1262, 1989. 18. Zincke, H., Bermtralh, E. J., Blute, M. L., Myers, R. P., Barrett, D. M., Liebe;, M. M., Martin, S. K. and Oesterling. J. E.: Radical prostatectomy for clinically localized prostate cancer: long-term results of 1,143 patients from a single institution. J. Clin. Oncol., 1 2 2254,1994. 19. Hanks, G. E., Asbell, S., Krall, J. M., Perez, C. A,, Doggett, S., Rubin, P., Sause, W. and Pilepich, M. V.: Outcome for lymph node dissection negative T-lb, T-2 (A-2, B) prostate cancer treated with external beam radiation therapy in RTOG 77-06. Int. J. Rad. Oncol. Biol. Phys., 21: 1099,1991. 20. Zietman, A. L., Coen, J. J., Dallow, K. C. and Shipley, W. U.: The treatment of prostate cancer by conventiona1 radiation therapy: an analysis of long-term outcome. Int. J. Rad. Oncol. Biol. Phys., 3 2 287,1995. 21. Schellhammer, P.J.,el-Mahdi, A. M.,Wright, G. L., Jr., Kolm, P. and Ragle, R.: Prostate-specific antigen to determine progression-free survival after radiation therapy for localized carcinoma of prostate. Urology, 4 2 13, 1993. EDITORIAL COMMENT

This well written article has several distinguishing features to its credit, including long followup (minimum 7 years), acceptable loss to followup incidence (11%), surgical staging by node dissection in all patients (to their knowledge the authors report the largest such surgically staged series), and inclusion of information on stage; grade (56% cases were retrieved for contemporary grading by 1 pathologist);co-morbid mortality, and clinical, biopsy and PSA progression (available in 79.1% of patients) with definitions of PSA progression criteria. Progression curves were constructed to include all failures, that is patients with clinical failure and those manifesting biochemical failure. alone, even if only 1elevation occurred that was unconfirmed by a second repeat elevation. Nevertheless, the calculated curves are likely overestimations because early followup was in the pre-PSA era so that the PSA failure date on the curve often did not represent the initial PSA failure, and PSA often was not measured in patients dying or lost to followup. These observations render the comparison with other series, suffering from similar retrospective variations and having different end point definitions, followup intervals and so forth, less meaningful. The prognostic significance of post-irradiation PSA nadir is well recognized but consensus has not been reached on a uniform, absolute numerical value.l.2 A consensus conference in September 1996 defined failure of definitive radiation therapy (PSA failure) as 3 subsequent elevations greater than the PSA nadir level with failure date identified as the halfway point between the date of nadir and the data of the first elevation.> This definition will facilitate comparisons among irradiation series, which

PROSTATE CANCER PATIENTS TREATED WITH EXTERNAL BEAM RADIATION THERAPY

1759

to date have used different PSA nadir levels as a failure end point, A shift from a treatment paradigm directed at total tumor burden and will also identify as failures cases with nadir levels of less than eradication to one emphasizing tumor control has been proposed and 1.0 ne./ml. and even less than 0.5 n-d m l . with successive incremen- is worth reviewing.5 Such a paradigm may be particularly applicable increases in PSA. in the population of elderly men whose ebbing years are being measIt may be that a significant proportion of patients with low PSA ured and valued in serum PSA units. nadirs after irradiation harbor tumors of minimal biological activity. Their response to irradiation, identified by the low nadir PSA, could Paul F. Schellhammer be selecting a favorable phenotype. Thus the outcome in this group Department of Urology with regard to disease-free rate may be due to the identification of Eastern Virginia Medical S c h l favorable subgroups as much as the treatment given.4 Norfolk, Virginia A dilemma persists for clinicians treating carcinoma of the prostate with surgery or radiation therapy. This dilemma revolves about 1. Schellhammer, P. F.:The role of PSA in the radiotherapy of reconciling end points critical for clinical trial assessment of treatprostate cancer. Editorial. Oncology, 1 0 1158,1996. ment effect and end points meaningful to patient assessment of 2. Schellhammer, P. F.,el-Mahdi, A. M., Kuban, D. A. and Wright, treatment. A slowly increasing PSA alone, 10 years aRer treatment, G. L., Jr.: PSA following radiation therapy: prognosis by preindicates failure of disease eradication and will undoubtedly prompt treatment level and posttreatment nadir. Uml. Clin. N. h e r . , modifications (intensification of therapy, combinations, adjuvant In press. and neoadjuvant treatments) to increase the likelihood of permanent 3. ASTRO Consensus Conference, September 28, 1996.San Antodisease eradication in future trials, which are necessary strategies nio, Texas. for the man in his 50s (age range in this series was 50 to 81 years). 4. McLauehlin. P. W.. Sander. H. M. and Jiroutek. M. R.: Prostate However, the 77-year-old patient (average age a t diagnosis in this specific antigen following. prostate radiotherapy: how low can study was 66.7 years) maintained clinically symptom-free and you go? Editorial. J. Clin. Oncol., 14: 2889,1996. metastasis-free for the previous decade has likely been well served 5. Schipper, H., Goh, C. R. and Wang, T. L.: Shifting the cancer by a treatment that has resulted in delayed disease progression. paradigm: must we kill to cure? Editorial. J. Clin. Oncol., IS: Only quality of life studies and determination of ultimate cause of 801,1995. death will clarify the issue.