- Email: [email protected]
FAILURE AFTER PRIMARY RADIATION OR SURGERY FOR PROSTATE CANCER: DIFFERENCES IN RESPONSE TO ANDROGEN ABLATION
0022-5347/04/1722-0525/0 THE JOURNAL OF UROLOGY® Copyright © 2004 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 172, 525–528, August 2004 Printed in U.S.A.
DOI: 10.1097/01.ju.0000132412.74468.57
FAILURE AFTER PRIMARY RADIATION OR SURGERY FOR PROSTATE CANCER: DIFFERENCES IN RESPONSE TO ANDROGEN ABLATION GREGORY P. SWANSON, MARK RIGGS
AND
JOHN EARLE
From Cancer Care Northwest, Spokane, Washington (GPS), Scott and White Clinic, Temple, Texas (MR), and Mayo Clinic, Jacksonville, Florida (JE)
ABSTRACT
Purpose: Androgen ablation is the standard treatment for recurrent and metastatic prostate cancer. Surprisingly few studies have documented the specific results for local and distant failure in patients treated primarily with radiation or radical prostatectomy. We report the long-term outcome of a series of those patients. Materials and Methods: We followed until death 94 patients in whom primary radiation therapy failed and 67 in whom radical prostatectomy failed. All patients received androgen ablation. Results: Statistically (p ⫽ 0.04) more patients in the radiation group (78%) died of prostate cancer than in the radical prostatectomy group (63%). Of the radiation group with local failure alone 63%, died of prostate cancer at a median of 5.03 years. Of the surgery group with isolated local failure 50% died of cancer at a median of 9.83 years. Of the patients treated with radiation with distant metastasis 93% died of cancer with a median time to death of 2.34 years. Of the patients treated with surgery 69% died of prostate cancer at a median of 3.27 years. The differences in survival between the 2 groups was significant. Conclusions: This study is unique in providing followup until death of patients treated with radical prostatectomy and radiation who had clinical failure and were treated with androgen ablation. Compelling is the finding that survival after androgen ablation after surgical failure is superior to that for radiation. If confirmed, this would be a significant consideration for future studies of patients in whom primary therapy fails. KEY WORDS: androgens, prostatic neoplasms, radiotherapy, prostatectomy
Since 1940, androgen ablation has been the standard treatment for metastatic prostate cancer. Despite its widespread use, detailed evaluation of response to treatment for failure after surgery or radiation is lacking, which is due to the necessity of long-term followup to determine actual outcome of treatment. We report on a group of patients treated with surgery and radiation followed until death and evaluate response to salvage androgen ablation. MATERIALS AND METHODS
A total of 136 consecutive patients underwent primary radiation therapy with curative intent from July 1964 to March 1973 (only 2 patients were treated before 1966). Median patient age was 64 years (range 44 to 80). During the same period (January 1966 to December 1973) 175 patients underwent perineal prostatectomy (PP) (44) or radical retropubic prostatectomy (RRP) (131) as primary treatment for prostate cancer. Median patient age was 61 years (range 38 to 71). Choice of surgical approach was by physician preference. Twenty patients (RRP 15 and PP 5) also underwent orchiectomy at time of surgery for high risk findings. No patients treated with surgery received radiation. The treatment specifics have been previously published.1 All patients had histologically confirmed prostate adenocarcinoma with no evidence of metastatic disease on bone scan or other radiographs. The patients treated with radiation received 60 Gy to the prostate, except for 2 who received 65 Gy and 1 each who received 62.5, 67 and 70 Gy. For the radiation group median followup was 11.2 years (range 4 months to 32.4 years). For the patients still alive followup was from 22.9 to 32.4 years (median 25.6). For the patients Accepted for publication March 19, 2004.
treated with surgery median followup was 17 years (range 1 month to 29 years). For the surviving patients followup was from 22.5 to 29.3 years. The response to primary treatment for both groups of patients has been previously reported.2, 3 Toward the end of the study period followup was obtained primarily via correspondence. Cancer specific followup was requested but generally as the patients got older and more debilitated, this was not routinely performed. Failure was determined by palpation of the prostate fossa, a positive biopsy or radiographically, often in conjunction with an increased prostatic acid phosphatase. An increased prostate specific antigen (PSA) alone was not considered a criterion for failure in this study. Since routine exams (ie rectal) became less frequent with time, late failures would be underreported unless symptoms developed. This study represents those cases identified to have failed. Almost all (95%) of the patients have died, and so this is a complete accounting of outcome. No patient died of prostate cancer without having received androgen ablation. Survival curves are computed using the Kaplan-Meier product-limit methodology. Disease-free survival is computed by counting a subsequent recurrence or death as an event. Curves are compared using the log rank test with p ⬍0.05 considered statistically significant. RESULTS
Table 1 shows the pattern of initial failure for each of the treatment groups. Of the 136 patients receiving primary radiation 94 had recurrences. Of the 131 patients treated with radical prostatectomy 50 recurrence and of the 44 treated with perineal prostatectomy 17 had recurrence. The pattern of failure was similar for each group. Local failure
525
526
FAILURE AFTER PRIMARY RADIATION OR SURGERY FOR PROSTATE CANCER TABLE 1. Initial site of failure by treatment
Treatment (No. pts) Radiation (94) RRP (50) Perineal (17) Total (161) * Lung, supraclavicular lymph node.
No. Local (%)
No. Bone (%)
No. Both (%)
No. Other (%)*
49 (52) 25 (50) 7 (41)
35 (37) 16 (32) 7 (41)
6 (6) 7 (14) 2 (12)
4 (4) 2 (4) 1 (6)
81 (50)
58 (36)
15 (9)
7 (4)
was seen in 58%, 64% and 53% of the radiation, RRP and PP groups, respectively. Bone metastasis was detected in 43%, 46% and 53% of the radiation, RRP and PP groups, respectively. There was no difference in median patient age at the time of initial treatment (radiation 63 years, RRP 62 years and PP 63 years). The failing surgery and radiation cases were equal as to grade (p ⫽ 0.35). Of the radiation cases 10% were grade 1, 66% grade 2, 22% grade 3 and 2% grade 4. Of the surgery cases 4%, 75%, 21% and 0 were grades 1 to 4, respectively. All patients were treated with androgen ablation at some point after failure (including most with prior orchiectomy). Table 2 shows the survival for each group. Two patients treated with radiation did not have information on disease status at time of death and are counted as dying with progressive prostate cancer. Progressive cancer (dead with disease) is defined as active disease with some contribution to death. One patient in the radiation group and 1 in the RRP group are alive with active disease after attempt at salvage and are included in the dead with progression group. Dead with no evidence of disease (NED) indicates patients with control of cancer after hormonal salvage who died without obvious contribution from prostate cancer. The number of patients dying of prostate cancer was statistically significant (p ⫽ 0.04) between the radiation (78%) and surgery (63%) group. For patients treated with radiation with local failure alone median time to subsequent progression (ie failure of androgen ablation) was 2.59 years, median time until death was 5.03 years and a total of 63% died of disease. For the patients treated with surgery the same parameters were 3.95 years, 9.83 years and 50%, respectively. Similar findings were seen for metastasis. For the patients treated with radiation time to failure after treatment was 1.16 years and time to death was 2.34 years with 93% dying of active disease. For the patients treated with surgery these findings were 2.67 years, 3.27 years and 69%, respectively. These differences were statistically significant (for those dying after initial local failure or distant disease). Figures 1 and 2 show overall survival (OS) and disease-free survival (DFS). Of cases initially treated with radiation and then failing after treatment with androgen ablation only 20%
were progression-free at 5 years and 3% at 10 years. Survival was 36% at 5 years and 8% at 10 years. No patient survived to 15 years. For the patients initially treated with surgery progression-free survival was 43% at 5 years, 26% at 10 years, 12% at 15 years and 7% at 20 years. Survival was 57% at 5 years, 35% at 10 years and 6% at 20 years. The differences in disease-free and overall survival between patients treated with radiation and surgery was statistically significant. Table 3 shows the effect on type of initial failure and the resulting disease-free and overall survival after androgen ablation for the patients treated with radiation and surgery. Cases with the simultaneous finding of local and bone disease were classified as distant metastasis. For the patients treated with radiation, the site of initial failure (local vs distant) only made a difference in disease-free and overall survival at 5 years and by 10 years there was no difference. Of the patients treated with surgery those with local failure did much better than those with disseminated disease out to 10 years but out to 15 and 20 years the results were similar. In the local failure and distant failure cases progression-free survival was better in the surgery than in the radiation group to a similar degree. Of the initial group of 175 patients treated with surgery, 20 underwent orchiectomy at the time of surgery, and only 8 of them are in this cohort. All patients were treated with further androgen ablation at the time of progression. Removing those patients from analysis does not significantly change the results (for example, 5-year survival increases from 56.7% to 62.1% and 10-year from 35.4% to 39.2%). DISCUSSION
This appears to be the first published report of a difference in response to androgen ablation between patients treated with surgery or radiation. In addition it is one of the few reports that provides detailed data on the outcome of patients treated with androgen ablation after failure of potentially curative treatment. The available literature suffers from short followup, making the ultimate outcome uncertain. Most currently reported studies are PSA based and there is little information being collected on clinical failure.
TABLE 2. Modality and site of failure with subsequent outcome
Radiation: Local Bone Other Totals Surgery: Local Bone Other
No. Pts
No. Alive NED (%)
No. Dead NED (%)
No. Dead With Disease (%)
Median Yrs to Subsequent Failure (progression)
Median Yrs to Death
49 41 4
4 (8) 0 (0) 0 (0)
14 (29) 3 (7) 0 (0)
31 (63) 38 (93) 4 (100)
2.59 1.16 0.38
5.03 2.34 1.93
94
4 (4)
17 (18)
73 (78)
32 32 3
4 (13) 1 (3) 0 (0)
12 (38) 7 (22) 1 (33)
16 (50) 22 (69) 2 (67)
3.95 2.67 6.86
9.83 3.27 7.70
5 (7)
20 (30)
40 (63)
8 (10) 1 (1) 0 (0)
26 (32) 10 (14) 1 (14)
47 (58) 60 (82) 6 (86)
2.93 1.18 1.20
5.65 2.54 2.83
9 (6)
37 (23)
113 (70)
1.91
3.98
Totals 67 All pts (radiation and surgery): Local 81 Bone 73 Other 7 Totals
161
FAILURE AFTER PRIMARY RADIATION OR SURGERY FOR PROSTATE CANCER
FIG. 1. Survival after failure by initial treatment treated with androgen ablation.
FIG. 2. Progression-free survival after androgen ablation by treatment.
Local failure after radiation. From the Patterns of Care Studies for primary radiation median survival after local failure was 3.5 years (5-year survival 42%) for stage B cases and 2.8 years (5-year survival 22%) for stage C cases.4 It is unclear whether all patients received androgen ablation. In a series from Eastern Virginia evaluating failure after external beam or implant therapy for stage A–B cases with local failure median survival was 70 months.5 There was no statistical difference between initial stage A, B and C, and all patients received androgen ablation. Only half of the patients were followed until death and followup from time of failure was not specified. In another study of cases treated with external beam plus implant therapy and then failing locally those with stages A2–C1 were reported together and had a median cause specific survival of 11 years after local failure but no details were given.6 Overall survival and length of
527
followup from failure were not reported. Our series consisted of patients with stage B and C disease. There was no difference in response to androgen ablation between the 2 stages (not shown), and so we considered them together. We saw a median overall survival after local failure of 5 years and 5-year survival was 53%. It is difficult to compare the results due to incompleteness of followup in most studies. Overall, it appears that the median survival after local failure in patients treated with radiation is around 5 years. Distant failure after radiation. For patients treated with radiation who have bone metastasis the Patterns of Care Studies reported a median survival of 2.7 years for stage B and 1.8 years for stage C.4 For the Eastern Virginia study only the subgroup with local failure and distant failure was reported.5 Failure occurred at 1.8 years for stages A and B and at 2.5 years for stage C. Again, followup was limited. We saw a 16% 5-year survival with a median survival of 2.3 years. Overall, it appears that median survival after distant metastasis in radiation treated patients is about 28 months. Local failure after surgery. No information could be located on the specific outcome of androgen ablation after clinical local failure in patients treated with surgery. In our series median time to subsequent progression was 4 years and only 50% died of prostate cancer. The 5-year survival rate was 75% and 10-year rate was 53%. Distant failure after surgery. In a series of 315 patients with an increasing PSA after prostatectomy after exclusions 103 had metastatic disease.7 Survival rates were 43% and 13% at 5 and 10 years, respectively. Only 43% of patients died of cancer. The study was designed to look at time from PSA increase to development of metastatic disease, and so the specifics of followup from the time of metastatic disease are not given. It is not clear how many of the total patients died. The reported median time of death after development of prostate cancer was “slightly less than 5 years.” Most studies of patients with metastatic disease do not specify the initial treatment. Therefore, it is uncertain if there are other comparative studies. For our prostatectomy cases that failed distally median time to death was 3.3 years, 5-year survival was 40% and 10-year survival was 20%. Distant failure primary treatment not specified. There are many studies of patients with metastatic prostate cancer looking at different methods of androgen ablation. These studies contain a mixture of patients with no prior treatment, radiation or surgery. The 2 largest studies were done by the Southwest Oncology Group. The first study, comparing luprolide (L) to luprolide plus flutamide (L ⫹ F), had more than 300 patients in each arm.8 The DFS and OS was 13.9 and 28.3 months for luprolide alone and 16.5 and 35.6 months for luprolide plus flutamide, respectively. Only about half of the patients died at the time of the report. In a larger study almost 700 patients in each arm were treated with orchiectomy or orchiectomy plus flutamide.9 In the orchiectomy arm DFS was 18.6 months and OS was 29.9 months. For the combined arm DFS was 20.4 months and OS was 33.5 months. Median followup was 50 months but completeness of followup (ie number of surviving patients) was not stated. It appears then that the best results with modern androgen ablation for metastatic disease is a DFS of about 20 months and an OS of about 35 months. In our radiation cases failing distally disease-free survival was 13.9 months and overall survival was 28.1 months. In our patients treated with surgery, disease-free survival was 32 months and overall survival was 39.2 months. In a subsequent analysis of Southwest Oncology Group 8894, the orchiectomy ⫾ flutamide study, initial treatment was evaluated for effect on survival.10 Detailed survival results were not given, but patients who had undergone prior radical prostatectomy had a better survival outcome (HR 0.77) than those treated with radiation (HR 1.22) when compared to those who had no prior treatment. It was noted that
528
FAILURE AFTER PRIMARY RADIATION OR SURGERY FOR PROSTATE CANCER TABLE 3. Survival and disease-free survival by initial treatment and site of failure treated with androgen ablation Local
% Overall survival from initial failure (yrs): 5 10 15 20 % Disease-free survival from initial failure (yrs): 5 10 15 20
Distant Metastasis
Radiation
Surgery
Radiation
Surgery
53.2 7.6 0.0 0.0
75.0 52.5 12.3 8.2
16.3 7.0 0.0 0.0
40.0 20.0 14.3 3.6
32.6 2.7 0.0 0.0
53.6 31.0 10.3 5.2
5.0 2.5 0.0 0.0
33.3 21.2 12.1 4.0
patients who had previously received radiation had worse prognostic factors at the time of androgen ablation (higher median PSA, higher grade, more advanced age) than those who had undergone prior prostatectomy, and so the groups are not directly comparable. Still, the differences in results between cases failing after prostatectomy versus radiation is interesting. Surgery vs radiation. In our study we looked at a series of patients treated with 2 different modalities at the same institution but treated during the same timeframe. The groups are not directly comparable because of the selection factors that made the radiation cases ineligible for surgery. If everything else was equal, the type of primary treatment should not be an issue in cases that fail. Clearly this was the thinking in the large randomized studies done on metastatic disease, where the primary treatment was not mentioned in initial reports. In our series there was a significant difference in the salvage rate for failure after surgery compared to failure after radiation. This may be explained by the fact that local failures in an intact gland after radiation are likely to be bulkier than those detected in the prostate fossa after radical prostatectomy. At the time of initial treatment the patients treated with radiation had bulky, palpable cancers and it would take a significant amount of disease to detect clinical recurrence. If that is the case, then our results would indicate that response to androgen ablation is influenced by bulk of disease, since the patients treated with radiation did worse. This observation should not be an issue for metastatic disease cases in which we also found a significant difference between radiation and surgery. It is possible that the patients treated with radiation had undetected concomitant local failure and that the difference was again related to bulk of recurrent disease. Otherwise, unless there is a large difference in followup between the radiation and surgery populations, we would expect the bulk of distant disease to be about equal. If that is the case, then it would appear that prostate cancer after radiation failure is more resistant to androgen ablation than cancer after surgical failure. Given that radiation and androgen ablation cause apoptosis, it is intriguing to consider that cells resistant to one modality might be more resistant to the other due to common pathways. For example, it has been shown that p53 levels in patients with radiation failure increased from 8% to 54%.11 Bcl2, which is a mitochondrial anti-apoptotic protein, increased from 33% to 67%. A corollary to this is that prostate cancer that recurs after radiation is often a higher grade than the initial cancer. Higher grade cancers in general do not respond as well.
CONCLUSIONS
This appears to be the first study with followup until death of patients with prostate cancer and systemic metastasis after primary radiation and surgery. It appears that those treated with surgery have a better response to subsequent androgen ablation. This finding needs to be confirmed but until proven otherwise, competing arms in studies looking at androgen ablative treatment of prostate cancer recurrence should be balanced as to whether the primary treatment was surgery or radiation. REFERENCES
1. Swanson, G. P., Cupps, R. E., Utz, D. C., Ilstrup, D. M., Zincke, H. and Myers, R. P.: Definitive therapy for prostate carcinoma: Mayo Clinic results at 15 years after treatment. Br J Radiol, 67: 877, 1994 2. Swanson, C. P., Riggs, M. W., Earle, J. D. and Haddock, M. G.: Long-term follow-up of radical retropubic prostatectomy for prostate cancer. Eur Urol, 42: 212, 2002 3. Swanson, G. P., Riggs, M. W., Earle, J. D. and Haddock, M. G.: Long-term follow-up of radical retropubic prostatectomy for prostate cancer. Eur Urol, 42: 212, 2002 4. Hanks, G. E., Diamond, J. J., Krall, J. M., Martz, K. L. and Kramer, S.: A ten year follow-up of 682 patients treated for prostate cancer with radiation therapy in the United States. Int J Radiat Oncol Biol Phys, 13: 499, 1987 5. Schellhammer, P. F., Kuban, D. A. and El-Mahdi, A. M.: Treatment of clinical local failure after radiation therapy for prostate cancer. J Urol, 150: 1851, 1993 6. Eastham, J. A., Kattan, M. W., Groshen, S., Scardino, P. T., Rogers, E., Carlton, C. E., Jr. et al: Fifteen-year survival and recurrence rates after radiotherapy for localized prostate cancer. J Clin Oncol, 15: 3214, 1997 7. Pound, C. R., Partin, A. W., Eisenberger, M. A., Chan, D. W., Pearson, J. D. and Walsh, P. C.: Natural history of progression after PSA elevation following radical prostatectomy. JAMA, 281: 1591, 1999 8. Crawford, E. D., Eisenberger, M. A., McLeod, D. G., Spaulding, J. T., Benson, R., Dorr, F. A. et al: A controlled trial of leuprolide with and without flutamide in prostatic carcinoma. N Engl J Med, 321: 419, 1989 9. Eisenberger, M. A., Blumenstein, B. A., Crawford, E. D., Miller, G., McLeod, D. G., Loehrer, P. J. et al: Bilateral orchiectomy with or without flutamide for metastatic prostate cancer. N Engl J Med, 339: 1036, 1998 10. Thompson, I. M., Tangen, C., Basler, J. and Crawford, E. D.: Impact of previous local treatment for prostate cancer on subsequent metastatic disease. J Urol, 168: 1008, 2002 11. Grossfeld, G. D., Olumi, A. F., Connolly, J. A., Chew, K., Gibney, J., Bhargava, V. et al: Locally recurrent prostate tumors following either radiation therapy or radical prostatectomy have changes in Ki-67 labeling index, p53 and bcl-2 immunoreactivity. J Urol, 159: 1437, 1998
Vol. 172, 525–528, August 2004 Printed in U.S.A.
DOI: 10.1097/01.ju.0000132412.74468.57
FAILURE AFTER PRIMARY RADIATION OR SURGERY FOR PROSTATE CANCER: DIFFERENCES IN RESPONSE TO ANDROGEN ABLATION GREGORY P. SWANSON, MARK RIGGS
AND
JOHN EARLE
From Cancer Care Northwest, Spokane, Washington (GPS), Scott and White Clinic, Temple, Texas (MR), and Mayo Clinic, Jacksonville, Florida (JE)
ABSTRACT
Purpose: Androgen ablation is the standard treatment for recurrent and metastatic prostate cancer. Surprisingly few studies have documented the specific results for local and distant failure in patients treated primarily with radiation or radical prostatectomy. We report the long-term outcome of a series of those patients. Materials and Methods: We followed until death 94 patients in whom primary radiation therapy failed and 67 in whom radical prostatectomy failed. All patients received androgen ablation. Results: Statistically (p ⫽ 0.04) more patients in the radiation group (78%) died of prostate cancer than in the radical prostatectomy group (63%). Of the radiation group with local failure alone 63%, died of prostate cancer at a median of 5.03 years. Of the surgery group with isolated local failure 50% died of cancer at a median of 9.83 years. Of the patients treated with radiation with distant metastasis 93% died of cancer with a median time to death of 2.34 years. Of the patients treated with surgery 69% died of prostate cancer at a median of 3.27 years. The differences in survival between the 2 groups was significant. Conclusions: This study is unique in providing followup until death of patients treated with radical prostatectomy and radiation who had clinical failure and were treated with androgen ablation. Compelling is the finding that survival after androgen ablation after surgical failure is superior to that for radiation. If confirmed, this would be a significant consideration for future studies of patients in whom primary therapy fails. KEY WORDS: androgens, prostatic neoplasms, radiotherapy, prostatectomy
Since 1940, androgen ablation has been the standard treatment for metastatic prostate cancer. Despite its widespread use, detailed evaluation of response to treatment for failure after surgery or radiation is lacking, which is due to the necessity of long-term followup to determine actual outcome of treatment. We report on a group of patients treated with surgery and radiation followed until death and evaluate response to salvage androgen ablation. MATERIALS AND METHODS
A total of 136 consecutive patients underwent primary radiation therapy with curative intent from July 1964 to March 1973 (only 2 patients were treated before 1966). Median patient age was 64 years (range 44 to 80). During the same period (January 1966 to December 1973) 175 patients underwent perineal prostatectomy (PP) (44) or radical retropubic prostatectomy (RRP) (131) as primary treatment for prostate cancer. Median patient age was 61 years (range 38 to 71). Choice of surgical approach was by physician preference. Twenty patients (RRP 15 and PP 5) also underwent orchiectomy at time of surgery for high risk findings. No patients treated with surgery received radiation. The treatment specifics have been previously published.1 All patients had histologically confirmed prostate adenocarcinoma with no evidence of metastatic disease on bone scan or other radiographs. The patients treated with radiation received 60 Gy to the prostate, except for 2 who received 65 Gy and 1 each who received 62.5, 67 and 70 Gy. For the radiation group median followup was 11.2 years (range 4 months to 32.4 years). For the patients still alive followup was from 22.9 to 32.4 years (median 25.6). For the patients Accepted for publication March 19, 2004.
treated with surgery median followup was 17 years (range 1 month to 29 years). For the surviving patients followup was from 22.5 to 29.3 years. The response to primary treatment for both groups of patients has been previously reported.2, 3 Toward the end of the study period followup was obtained primarily via correspondence. Cancer specific followup was requested but generally as the patients got older and more debilitated, this was not routinely performed. Failure was determined by palpation of the prostate fossa, a positive biopsy or radiographically, often in conjunction with an increased prostatic acid phosphatase. An increased prostate specific antigen (PSA) alone was not considered a criterion for failure in this study. Since routine exams (ie rectal) became less frequent with time, late failures would be underreported unless symptoms developed. This study represents those cases identified to have failed. Almost all (95%) of the patients have died, and so this is a complete accounting of outcome. No patient died of prostate cancer without having received androgen ablation. Survival curves are computed using the Kaplan-Meier product-limit methodology. Disease-free survival is computed by counting a subsequent recurrence or death as an event. Curves are compared using the log rank test with p ⬍0.05 considered statistically significant. RESULTS
Table 1 shows the pattern of initial failure for each of the treatment groups. Of the 136 patients receiving primary radiation 94 had recurrences. Of the 131 patients treated with radical prostatectomy 50 recurrence and of the 44 treated with perineal prostatectomy 17 had recurrence. The pattern of failure was similar for each group. Local failure
525
526
FAILURE AFTER PRIMARY RADIATION OR SURGERY FOR PROSTATE CANCER TABLE 1. Initial site of failure by treatment
Treatment (No. pts) Radiation (94) RRP (50) Perineal (17) Total (161) * Lung, supraclavicular lymph node.
No. Local (%)
No. Bone (%)
No. Both (%)
No. Other (%)*
49 (52) 25 (50) 7 (41)
35 (37) 16 (32) 7 (41)
6 (6) 7 (14) 2 (12)
4 (4) 2 (4) 1 (6)
81 (50)
58 (36)
15 (9)
7 (4)
was seen in 58%, 64% and 53% of the radiation, RRP and PP groups, respectively. Bone metastasis was detected in 43%, 46% and 53% of the radiation, RRP and PP groups, respectively. There was no difference in median patient age at the time of initial treatment (radiation 63 years, RRP 62 years and PP 63 years). The failing surgery and radiation cases were equal as to grade (p ⫽ 0.35). Of the radiation cases 10% were grade 1, 66% grade 2, 22% grade 3 and 2% grade 4. Of the surgery cases 4%, 75%, 21% and 0 were grades 1 to 4, respectively. All patients were treated with androgen ablation at some point after failure (including most with prior orchiectomy). Table 2 shows the survival for each group. Two patients treated with radiation did not have information on disease status at time of death and are counted as dying with progressive prostate cancer. Progressive cancer (dead with disease) is defined as active disease with some contribution to death. One patient in the radiation group and 1 in the RRP group are alive with active disease after attempt at salvage and are included in the dead with progression group. Dead with no evidence of disease (NED) indicates patients with control of cancer after hormonal salvage who died without obvious contribution from prostate cancer. The number of patients dying of prostate cancer was statistically significant (p ⫽ 0.04) between the radiation (78%) and surgery (63%) group. For patients treated with radiation with local failure alone median time to subsequent progression (ie failure of androgen ablation) was 2.59 years, median time until death was 5.03 years and a total of 63% died of disease. For the patients treated with surgery the same parameters were 3.95 years, 9.83 years and 50%, respectively. Similar findings were seen for metastasis. For the patients treated with radiation time to failure after treatment was 1.16 years and time to death was 2.34 years with 93% dying of active disease. For the patients treated with surgery these findings were 2.67 years, 3.27 years and 69%, respectively. These differences were statistically significant (for those dying after initial local failure or distant disease). Figures 1 and 2 show overall survival (OS) and disease-free survival (DFS). Of cases initially treated with radiation and then failing after treatment with androgen ablation only 20%
were progression-free at 5 years and 3% at 10 years. Survival was 36% at 5 years and 8% at 10 years. No patient survived to 15 years. For the patients initially treated with surgery progression-free survival was 43% at 5 years, 26% at 10 years, 12% at 15 years and 7% at 20 years. Survival was 57% at 5 years, 35% at 10 years and 6% at 20 years. The differences in disease-free and overall survival between patients treated with radiation and surgery was statistically significant. Table 3 shows the effect on type of initial failure and the resulting disease-free and overall survival after androgen ablation for the patients treated with radiation and surgery. Cases with the simultaneous finding of local and bone disease were classified as distant metastasis. For the patients treated with radiation, the site of initial failure (local vs distant) only made a difference in disease-free and overall survival at 5 years and by 10 years there was no difference. Of the patients treated with surgery those with local failure did much better than those with disseminated disease out to 10 years but out to 15 and 20 years the results were similar. In the local failure and distant failure cases progression-free survival was better in the surgery than in the radiation group to a similar degree. Of the initial group of 175 patients treated with surgery, 20 underwent orchiectomy at the time of surgery, and only 8 of them are in this cohort. All patients were treated with further androgen ablation at the time of progression. Removing those patients from analysis does not significantly change the results (for example, 5-year survival increases from 56.7% to 62.1% and 10-year from 35.4% to 39.2%). DISCUSSION
This appears to be the first published report of a difference in response to androgen ablation between patients treated with surgery or radiation. In addition it is one of the few reports that provides detailed data on the outcome of patients treated with androgen ablation after failure of potentially curative treatment. The available literature suffers from short followup, making the ultimate outcome uncertain. Most currently reported studies are PSA based and there is little information being collected on clinical failure.
TABLE 2. Modality and site of failure with subsequent outcome
Radiation: Local Bone Other Totals Surgery: Local Bone Other
No. Pts
No. Alive NED (%)
No. Dead NED (%)
No. Dead With Disease (%)
Median Yrs to Subsequent Failure (progression)
Median Yrs to Death
49 41 4
4 (8) 0 (0) 0 (0)
14 (29) 3 (7) 0 (0)
31 (63) 38 (93) 4 (100)
2.59 1.16 0.38
5.03 2.34 1.93
94
4 (4)
17 (18)
73 (78)
32 32 3
4 (13) 1 (3) 0 (0)
12 (38) 7 (22) 1 (33)
16 (50) 22 (69) 2 (67)
3.95 2.67 6.86
9.83 3.27 7.70
5 (7)
20 (30)
40 (63)
8 (10) 1 (1) 0 (0)
26 (32) 10 (14) 1 (14)
47 (58) 60 (82) 6 (86)
2.93 1.18 1.20
5.65 2.54 2.83
9 (6)
37 (23)
113 (70)
1.91
3.98
Totals 67 All pts (radiation and surgery): Local 81 Bone 73 Other 7 Totals
161
FAILURE AFTER PRIMARY RADIATION OR SURGERY FOR PROSTATE CANCER
FIG. 1. Survival after failure by initial treatment treated with androgen ablation.
FIG. 2. Progression-free survival after androgen ablation by treatment.
Local failure after radiation. From the Patterns of Care Studies for primary radiation median survival after local failure was 3.5 years (5-year survival 42%) for stage B cases and 2.8 years (5-year survival 22%) for stage C cases.4 It is unclear whether all patients received androgen ablation. In a series from Eastern Virginia evaluating failure after external beam or implant therapy for stage A–B cases with local failure median survival was 70 months.5 There was no statistical difference between initial stage A, B and C, and all patients received androgen ablation. Only half of the patients were followed until death and followup from time of failure was not specified. In another study of cases treated with external beam plus implant therapy and then failing locally those with stages A2–C1 were reported together and had a median cause specific survival of 11 years after local failure but no details were given.6 Overall survival and length of
527
followup from failure were not reported. Our series consisted of patients with stage B and C disease. There was no difference in response to androgen ablation between the 2 stages (not shown), and so we considered them together. We saw a median overall survival after local failure of 5 years and 5-year survival was 53%. It is difficult to compare the results due to incompleteness of followup in most studies. Overall, it appears that the median survival after local failure in patients treated with radiation is around 5 years. Distant failure after radiation. For patients treated with radiation who have bone metastasis the Patterns of Care Studies reported a median survival of 2.7 years for stage B and 1.8 years for stage C.4 For the Eastern Virginia study only the subgroup with local failure and distant failure was reported.5 Failure occurred at 1.8 years for stages A and B and at 2.5 years for stage C. Again, followup was limited. We saw a 16% 5-year survival with a median survival of 2.3 years. Overall, it appears that median survival after distant metastasis in radiation treated patients is about 28 months. Local failure after surgery. No information could be located on the specific outcome of androgen ablation after clinical local failure in patients treated with surgery. In our series median time to subsequent progression was 4 years and only 50% died of prostate cancer. The 5-year survival rate was 75% and 10-year rate was 53%. Distant failure after surgery. In a series of 315 patients with an increasing PSA after prostatectomy after exclusions 103 had metastatic disease.7 Survival rates were 43% and 13% at 5 and 10 years, respectively. Only 43% of patients died of cancer. The study was designed to look at time from PSA increase to development of metastatic disease, and so the specifics of followup from the time of metastatic disease are not given. It is not clear how many of the total patients died. The reported median time of death after development of prostate cancer was “slightly less than 5 years.” Most studies of patients with metastatic disease do not specify the initial treatment. Therefore, it is uncertain if there are other comparative studies. For our prostatectomy cases that failed distally median time to death was 3.3 years, 5-year survival was 40% and 10-year survival was 20%. Distant failure primary treatment not specified. There are many studies of patients with metastatic prostate cancer looking at different methods of androgen ablation. These studies contain a mixture of patients with no prior treatment, radiation or surgery. The 2 largest studies were done by the Southwest Oncology Group. The first study, comparing luprolide (L) to luprolide plus flutamide (L ⫹ F), had more than 300 patients in each arm.8 The DFS and OS was 13.9 and 28.3 months for luprolide alone and 16.5 and 35.6 months for luprolide plus flutamide, respectively. Only about half of the patients died at the time of the report. In a larger study almost 700 patients in each arm were treated with orchiectomy or orchiectomy plus flutamide.9 In the orchiectomy arm DFS was 18.6 months and OS was 29.9 months. For the combined arm DFS was 20.4 months and OS was 33.5 months. Median followup was 50 months but completeness of followup (ie number of surviving patients) was not stated. It appears then that the best results with modern androgen ablation for metastatic disease is a DFS of about 20 months and an OS of about 35 months. In our radiation cases failing distally disease-free survival was 13.9 months and overall survival was 28.1 months. In our patients treated with surgery, disease-free survival was 32 months and overall survival was 39.2 months. In a subsequent analysis of Southwest Oncology Group 8894, the orchiectomy ⫾ flutamide study, initial treatment was evaluated for effect on survival.10 Detailed survival results were not given, but patients who had undergone prior radical prostatectomy had a better survival outcome (HR 0.77) than those treated with radiation (HR 1.22) when compared to those who had no prior treatment. It was noted that
528
FAILURE AFTER PRIMARY RADIATION OR SURGERY FOR PROSTATE CANCER TABLE 3. Survival and disease-free survival by initial treatment and site of failure treated with androgen ablation Local
% Overall survival from initial failure (yrs): 5 10 15 20 % Disease-free survival from initial failure (yrs): 5 10 15 20
Distant Metastasis
Radiation
Surgery
Radiation
Surgery
53.2 7.6 0.0 0.0
75.0 52.5 12.3 8.2
16.3 7.0 0.0 0.0
40.0 20.0 14.3 3.6
32.6 2.7 0.0 0.0
53.6 31.0 10.3 5.2
5.0 2.5 0.0 0.0
33.3 21.2 12.1 4.0
patients who had previously received radiation had worse prognostic factors at the time of androgen ablation (higher median PSA, higher grade, more advanced age) than those who had undergone prior prostatectomy, and so the groups are not directly comparable. Still, the differences in results between cases failing after prostatectomy versus radiation is interesting. Surgery vs radiation. In our study we looked at a series of patients treated with 2 different modalities at the same institution but treated during the same timeframe. The groups are not directly comparable because of the selection factors that made the radiation cases ineligible for surgery. If everything else was equal, the type of primary treatment should not be an issue in cases that fail. Clearly this was the thinking in the large randomized studies done on metastatic disease, where the primary treatment was not mentioned in initial reports. In our series there was a significant difference in the salvage rate for failure after surgery compared to failure after radiation. This may be explained by the fact that local failures in an intact gland after radiation are likely to be bulkier than those detected in the prostate fossa after radical prostatectomy. At the time of initial treatment the patients treated with radiation had bulky, palpable cancers and it would take a significant amount of disease to detect clinical recurrence. If that is the case, then our results would indicate that response to androgen ablation is influenced by bulk of disease, since the patients treated with radiation did worse. This observation should not be an issue for metastatic disease cases in which we also found a significant difference between radiation and surgery. It is possible that the patients treated with radiation had undetected concomitant local failure and that the difference was again related to bulk of recurrent disease. Otherwise, unless there is a large difference in followup between the radiation and surgery populations, we would expect the bulk of distant disease to be about equal. If that is the case, then it would appear that prostate cancer after radiation failure is more resistant to androgen ablation than cancer after surgical failure. Given that radiation and androgen ablation cause apoptosis, it is intriguing to consider that cells resistant to one modality might be more resistant to the other due to common pathways. For example, it has been shown that p53 levels in patients with radiation failure increased from 8% to 54%.11 Bcl2, which is a mitochondrial anti-apoptotic protein, increased from 33% to 67%. A corollary to this is that prostate cancer that recurs after radiation is often a higher grade than the initial cancer. Higher grade cancers in general do not respond as well.
CONCLUSIONS
This appears to be the first study with followup until death of patients with prostate cancer and systemic metastasis after primary radiation and surgery. It appears that those treated with surgery have a better response to subsequent androgen ablation. This finding needs to be confirmed but until proven otherwise, competing arms in studies looking at androgen ablative treatment of prostate cancer recurrence should be balanced as to whether the primary treatment was surgery or radiation. REFERENCES
1. Swanson, G. P., Cupps, R. E., Utz, D. C., Ilstrup, D. M., Zincke, H. and Myers, R. P.: Definitive therapy for prostate carcinoma: Mayo Clinic results at 15 years after treatment. Br J Radiol, 67: 877, 1994 2. Swanson, C. P., Riggs, M. W., Earle, J. D. and Haddock, M. G.: Long-term follow-up of radical retropubic prostatectomy for prostate cancer. Eur Urol, 42: 212, 2002 3. Swanson, G. P., Riggs, M. W., Earle, J. D. and Haddock, M. G.: Long-term follow-up of radical retropubic prostatectomy for prostate cancer. Eur Urol, 42: 212, 2002 4. Hanks, G. E., Diamond, J. J., Krall, J. M., Martz, K. L. and Kramer, S.: A ten year follow-up of 682 patients treated for prostate cancer with radiation therapy in the United States. Int J Radiat Oncol Biol Phys, 13: 499, 1987 5. Schellhammer, P. F., Kuban, D. A. and El-Mahdi, A. M.: Treatment of clinical local failure after radiation therapy for prostate cancer. J Urol, 150: 1851, 1993 6. Eastham, J. A., Kattan, M. W., Groshen, S., Scardino, P. T., Rogers, E., Carlton, C. E., Jr. et al: Fifteen-year survival and recurrence rates after radiotherapy for localized prostate cancer. J Clin Oncol, 15: 3214, 1997 7. Pound, C. R., Partin, A. W., Eisenberger, M. A., Chan, D. W., Pearson, J. D. and Walsh, P. C.: Natural history of progression after PSA elevation following radical prostatectomy. JAMA, 281: 1591, 1999 8. Crawford, E. D., Eisenberger, M. A., McLeod, D. G., Spaulding, J. T., Benson, R., Dorr, F. A. et al: A controlled trial of leuprolide with and without flutamide in prostatic carcinoma. N Engl J Med, 321: 419, 1989 9. Eisenberger, M. A., Blumenstein, B. A., Crawford, E. D., Miller, G., McLeod, D. G., Loehrer, P. J. et al: Bilateral orchiectomy with or without flutamide for metastatic prostate cancer. N Engl J Med, 339: 1036, 1998 10. Thompson, I. M., Tangen, C., Basler, J. and Crawford, E. D.: Impact of previous local treatment for prostate cancer on subsequent metastatic disease. J Urol, 168: 1008, 2002 11. Grossfeld, G. D., Olumi, A. F., Connolly, J. A., Chew, K., Gibney, J., Bhargava, V. et al: Locally recurrent prostate tumors following either radiation therapy or radical prostatectomy have changes in Ki-67 labeling index, p53 and bcl-2 immunoreactivity. J Urol, 159: 1437, 1998