Dr. Johnstone and Dr. Assikis are Georgia Cancer Coalition Distinguished Cancer Scholars and were supported in part by the Georgia Cancer Coalition.
Radical prostatectomy for clinical T4 prostate cancer
Article first published online: 12 MAY 2006
Copyright © 2006 American Cancer Society
Volume 106, Issue 12, pages 2603–2609, 15 June 2006
How to Cite
Johnstone, P. A. S., Ward, K. C., Goodman, M., Assikis, V. and Petros, J. A. (2006), Radical prostatectomy for clinical T4 prostate cancer. Cancer, 106: 2603–2609. doi: 10.1002/cncr.21926
- Issue published online: 1 JUN 2006
- Article first published online: 12 MAY 2006
- Manuscript Accepted: 31 JAN 2006
- Manuscript Revised: 15 DEC 2005
- Manuscript Received: 12 OCT 2005
- National Center on Minority Health and Health Disparities. Grant Number: 5P60-MD000525
- Georgia Cancer Coalition
- hormone therapy;
- prostate neoplasms;
- Surveillance Epidemiology and End Results
Occasionally, patients with clinical T4 (cT4) prostate cancer undergo surgery. Published data on outcomes after radical prostatectomy (RP) in patients with such advanced stage disease and on the impact of adjuvant radiation therapy (RT) and hormone therapy (HT) are nonexistent.
Data from the Surveillance Epidemiology and End Results (SEER) data base were reviewed for the 7-year period from 1995 to 2001. Specifically, data were analyzed for 1093 patients with cT4, lymph node-negative or lymph node-positive, M0 prostate cancer without distant lymph node involvement or a history of other cancer. Using follow-up data through 2002, postdiagnosis survival was examined in 5 treatment groups: radical prostatectomy (RP) either alone or in combination with other therapy, radiation therapy (RT) alone, hormone therapy (HT) alone, RT plus HT, and no treatment (NT). All results were expressed as 1-year, 3-year, and 5-year observed survival and corresponding relative survival. Mortality across treatment categories was compared by using a Cox proportionate hazards model controlling for age, year of diagnosis, race, tumor grade, regional lymph node involvement, clinical tumor extension, and SEER registry.
Observed and relative survival rates were lowest among patients who received NT and highest among patients who underwent RP. Adding adjuvant RT or HT to RP conferred no survival benefit. Multivariate survival analyses revealed a significant increase in mortality among HT-only patients and among patients who received NT compared with patients who underwent RP. The differences in survival among treatment types were most pronounced in a relatively small group of patients who had positive regional lymph node extension. In all other patients, the results suggested a modest (but not significant) improvement in survival after RT plus HT.
SEER data revealed that patients who underwent RP for cT4 prostate cancer had increased survival compared with patients who received RT alone or HT alone and had a survival comparable to that of patients who received RT plus HT. The benefit of RP appears to be limited to a relatively small subset of patients who have regional lymph node extension. Cancer 2006. © 2006 American Cancer Society.
Patients with clinical T4 (cT4) prostate cancer undergo radical prostatectomy (RP) infrequently. Although, generally, it is not considered effective in rendering a cure, surgery sometimes is performed either to enhance local control or to “debulk” the lesion. To our knowledge, published data regarding the outcomes of such RP and comparisons with other treatments, such as radiation therapy (RT) and hormone therapy (HT), are nonexistent.
There is a lack of information regarding treatment outcomes in patients with cT4 prostate cancer, because newly diagnosed cT4 cancers are relatively rare; consequently, very few institutions have sufficient numbers of patients with cT4 disease who undergo surgery to allow comparisons with other treatment categories. In this report, we present our analyses of the data from the Surveillance, Epidemiology, and End Results (SEER) data base, which represents approximately 26% of the U.S. population and permits more extensive multivariate analyses because of its large sample size. SEER registries collect detailed information regarding the extent of the disease and the first course of cancer-directed therapy. All data coding used by SEER is set up by a joint task force, which includes representatives from SEER, the American College of Surgeons, the Centers for Disease Control and Prevention, the North American Association of Central Cancer Registries, the National Cancer Registrars Association, and the American Joint Committee on Cancer. The objective of the joint task force is to provide a single, uniform set of codes and rules for coding to meet the needs of all participants in the standard-setting process.1 The objective of the current analysis was to compare the survival of patients with cT4 prostate cancer who underwent RP with the survival of patients who received other types of treatment or who received no treatment at all.
MATERIALS AND METHODS
Using SEER data, we selected all eligible patients with prostate cancer based on the following inclusion criteria: 1) invasive disease, 2) cT4 prostate cancer, 3) diagnosis between 1995 and 2001, 4) no clinical evidence of distant lymph node involvement, and 5) no other history of malignancy. Patients who were reported exclusively based on death certificates or autopsy findings were excluded. Although 97.8% of patients had their disease confirmed microscopically, patients with only clinical diagnoses were not excluded. cT4 was defined as tumor extension or fixation to any of the following adjacent structures other than the seminal vesicles: retrovesical fascia; bladder, not otherwise specified (NOS); urether; fixation, NOS; bladder neck; rectum; external sphincter; levator muscles; skeletal muscles, NOS; pelvic wall; or pelvic bone. The definition of RP included patients who underwent any of the following procedures: RP, total prostatectomy, or cystoprostatectomy. Patients who underwent exploratory surgery only or lymphadenectomy only and patients whose documented surgical procedures were limited to biopsy and/or transurethral resection of the prostate were excluded from the RP category. A decision to limit our analyses to the post-1994 data was based on the fact that, during earlier years, staging information in the SEER data base did not distinguish between the clinical stage and the pathologic stage.
Using follow-up data through 2002, survival was examined in the following 5 treatment groups: RP (alone or in combination with other therapy), RT alone, HT alone (defined as endocrine surgery/radiation or hormone medication), combined RT and HT, or no therapy at all (NT). For each of these 5 treatment groups, we calculated 1-year, 3-year, and 5-year observed survival (OS), which was defined as the proportion of patients who survived beyond a certain interval after the diagnosis, and the corresponding relative survival (RS), which was defined as the OS among cancer patients divided by the expected survival in the general population with the same age, gender, and race characteristics.
Multivariate analyses were conducted by using Cox proportional hazards models that, in addition to age and race, controlled for year of diagnosis, extension to regional lymph nodes, tumor grade, clinical tumor extension, and SEER registry. Clinical tumor extension differentiated extension or fixation to the pelvic wall or pelvic bone from the remaining structures, meeting the definition of cT4. SEER registry information was included to control for regional variation in treatment patterns. The multivariate analyses included only patients who had documented race and grade and compared survival across treatment categories using 3 versions of the Cox proportional hazard model. Model I included 5 treatment categories: RP (alone or in combination with other therapy), RT alone, HT alone, combined RT and HT, and NT, with “RP” used as the reference group. Model II included 6 treatment categories: RP alone, combined RP plus RT (and/or HT), RT alone, HT alone, combined RT and HT, and NT, with “RP alone” used as the reference group. Model III was the same as Model II but included only patients who had their regional lymph node status documented as either positive or negative. The results of the multivariate survival analyses are presented as hazards ratios (HRs) with the corresponding 95% confidence intervals (95% CIs). All models were examined for interaction between treatment and other patient characteristics. The analyses were performed using the software programs SPSS for Windows (version 11.5; LEAD Technologies, Inc., Chicago IL), SAS 9.1.3 (SAS Institute Inc., Cary, NC), and SEER*Stat (National Cancer Institute, Bethesda, MD).
In total, 1093 patients with cT4 prostate cancer met the inclusion criteria for this analysis; of these, only 72 patients (6.6%) underwent RP. Among those, 50 patients underwent surgical treatment alone, 13 patients underwent RP in combination with HT, and 9 patients underwent surgery in combination with either RT (n = 4) or combined RT and HT (n = 5). The patient characteristics are presented in Table 1. The majority of patients were Caucasians (74.7%), and most (79.9%) were age ≥65 years.
|Patient Characteristics||Overall||RP||RT||HT||RT and HT||NT|
|Age at diagnosis, y|
|Extension to regional lymph nodes|
|Clinical tumor extension|
|Pelvic wall or bone||135||12.4||6||4.4||10||7.4||53||39.3||44||32.6||22||16.3|
Table 1 shows that 14.5% of patients age <65 years underwent RP either alone or in combination with RT and/or HT. This proportion was higher than the corresponding percentages of 9.3% among patients ages 65 years and 74 years (P = .05) and 1.4% among patients age ≥75 years (P<.001). Conversely, older patients were more likely to receive NT (38.3%) compared with 17.4% of patients in the middle age group (P<.001) and 9.5% in the younger age group (P<.001).
In 57.9% of all patients with cT4 prostate cancer, the regional lymph node status was unknown. The majority of patients (98.1%) with unknown regional lymph node status were those whose treatment did not include surgery.
Table 2 shows that the OS and RS rates were lowest among patients who received NT and among those who received HT only. Patients who underwent prostatectomy and received combined RT plus HT appeared to have the highest survival rates, whereas the survival of patients who received RT alone was somewhat lower.
|Treatment Type||No. of Patients*||1 Year (%)||3 Years (%)||5 Years (%)|
|RT and HT||257||98.10||99.9†‡||82.70||93.60||71.10||87.2‡|
The multivariate Cox survival analysis controlled for age, year of diagnosis, race, tumor grade, extension to regional lymph nodes, clinical tumor extension, and SEER registry. Table 3 shows that patients who received HT alone (HR, 2.02; 95% CI, 1.21–3.37 for Model I; HR, 2.23; 95% CI, 1.18–4.21 for Model II) had significantly increased mortality compared with the reference groups (any RP for Model I and RP alone for Model II). The increase in mortality was particularly pronounced when RP was compared with NT (HR, 2.32; 95% CI, 1.37–3.94 for Model I; HR, 2.56; 95% CI, 1.34–4.90 for Model II). Mortality for patients who received RT alone seemed to be somewhat elevated compared with that for patients who underwent RP, although the difference was not statistically significant; whereas the mortality for patients who received combined RT and HT did not differ significantly from that for patients who underwent RP (either alone or in combination with other treatments) (Table 3).
|Model I*||Model II*||Model III†|
|Treatment Type||HR||95% CI||Treatment Type||HR||95% CI||Treatment Type||HR||95% CI|
|RP (alone, or in combination)||1.0||Reference||RP alone||1.0||Reference||RP alone||1.0||Reference|
|RP and (RT and/or HT)||1.31||0.5–3.43||RP and (RT and/or HT)||2.72||0.85–8.65|
|RT alone||1.49||0.81–2.74||RT alone||1.65||0.81–3.36||RT alone||3.04‡||1.15–8.08‡|
|HT alone||2.02‡||1.21–3.37‡||HT alone||2.23‡||1.18–4.21‡||HT alone||3.81‡||1.56–9.32‡|
|RT and HT||1.09||0.64–1.87||RT and HT||1.20||0.62–2.33||RT and HT||1.82||0.74–4.52|
When the analyses included only those patients with known regional lymph node status (Model III), the differences between RP alone and other treatment options became more evident. The HRs for HT alone and NT were 3.81 (95% CI, 1.56–9.32) and 4.25 (95% CI, 1.66–10.86), respectively. Moreover, there was a significant difference between RP alone and RT alone (HR, 3.04; 95% CI, 1.15–8.08), whereas the corresponding HR for patients who received combined RP and RT (and or HT) was elevated but not statistically significant (Table 3).
Table 4 examines the role of regional lymph node extension further by presenting analyses separately for patients with known negative lymph node status, patients with known positive lymph node status, and patients with unknown lymph node status. These analyses showed that the observed differences in survival among different treatment types were attributable primarily to a relatively small subset of patients who had positive regional lymph node extension (n = 77 patients), a group that represents only 7% of the total study population (n = 1093 patients) and 18% of patients with known regional lymph node status (n = 430 patients). In this group of patients the HRs for RT alone, HT alone, and NT were 18.85 (95% CI, 1.55–228.67), 14.53 (95% CI, 1.73–121.93), and 14. 56 (95% CI, 1.14–186.43), respectively. By contrast, in the groups with negative regional lymph node status and unknown regional lymph node status, the results suggested a modest (albeit not significant) improvement in survival associated with combined RT and HT compared with RP; whereas the corresponding HRs for HT alone and NT were elevated, but the confidence intervals also included unity.
|Negative Regional Lymph Node Extension (n = 353)||Positive Regional Lymph Node Extension (n = 77)||Unknown Regional Lymph Node Extension (n = 577)|
|Treatment Type||HR||95% CI||Treatment Type||HR||95% CI||Treatment Type||HR||95% CI|
|RP (alone, or in combination)||1.0||Reference||RP (alone or in combination)||1.0||Reference||RP (alone, or in combination)||1.0||Reference|
|RT alone||1.33||0.56–3.16||RT alone||18.85*||1.55–228.67*||RT alone||0.83||0.31–2.26|
|HT alone||1.81||0.80–4.12||HT alone||14.53*||1.73–121.93*||HT alone||1.33||0.58–3.08|
|RT and HT||0.81||0.37–1.77||RT and HT||7.37||0.76–71.66||RT and HT||0.71||0.29–1.77|
The current results shed light on an infrequent clinical situation: radical surgery for patients with cT4 prostate cancer. Of the hundreds of thousands of men who were diagnosed with prostate cancer during this era, only 72 men underwent RP for cT4 disease. Although there are hazards to inferring too much from these data, optimizing therapy for these patients may be more challenging than we believed previously. It is also important to point out that regional lymph node involvement appears to play critical role in predicting treatment-related survival.
It noteworthy that the observed results likely are affected by (perhaps inevitable) misclassification of clinical prostate cancer staging. This issue can be addressed by analyzing the agreement between clinical and pathologic stage among patients with cT4 prostate cancer who underwent RP. Of the 72 patients in this category, only 31 patients (43%) had pathologic Stage T4 disease, 24 patients (33%) had their pathologic stage downgraded to T3 or lower, and the pathologic stage remained unknown for 17 patients (40%). Clearly, this misclassification is not limited to surgical patients; rather, surgical patients are the only group for which the accuracy of clinical staging can be evaluated. Despite these concerns, it is worth remembering that clinical stage is the only information about disease progression that is available before the initial treatment decisions are made.
Approximately 10% of all new prostate cancers are diagnosed as locally advanced. Such patients are at higher risk for both local recurrence and systemic disease. Validated nomograms are used in clinical practice to guide clinicians: high-volume, palpable disease (≥cT2c), high Gleason score (≥8), and prostate-specific antigen (PSA) levels from ≥15 ng/mL to 20 ng/mL are the 3 variables that portend high risk.2 Cure rates in patients with ≥1 of such characteristics are <50% with any single modality therapy (RP or RT). Combined therapeutic approaches with local and systemic therapies, thus, have been explored in an effort to increase cure rates. Recently, a standard of care for these lesions has emerged that involves HT with definitive RT. These approaches are based on several large, randomized trials that have reported a benefit of combined RT plus HT over RT alone.3–5 Some patients benefit from neoadjuvant plus concurrent HT,3, 4 whereas adjuvant HT is sufficient for other patients.5
A number of Phase III RT trials have addressed the question of optimal integration of HT with RT: A European Organization for Research and Treatment of Cancer (EORTC)5 study of concurrent plus adjuvant HT for 3 years versus RT alone was the first to report not only a disease-free survival (DFS) benefit (34% at 5 years) but, more important, a 5-year overall survival benefit of 16%. Radiation Therapy Oncology Group (RTOG) Trial 86103 was a seminal study, and those investigators reported that the addition of HT for 4 months (2 months prior to and 2 months during RT) to RT significantly improved rates of local control, DFS, and cancer specific mortality by approximately 10% each. RTOG 85316, 7 addressed the role of adjuvant HT (starting in the last week of RT) in patients with locally advanced prostate cancer who received RT alone. At a median follow-up of 7.6 years, all clinical endpoints were improved in the combined RT and HT arm: Local failure improved by 15%, overall survival improved by 10%, and cancer specific mortality improved by 6%. RTOG 94135 showed that the combination of 4 months of HT (2 months prior to and 2 months during RT) along with whole-pelvis RT conferred a DFS (but not an overall survival) benefit in patients with locally advanced prostate cancer and a ≥15% risk of clinically positive pelvic lymph nodes. Finally, D'Amico et al. reported on a Phase III trial investigating the combination of 70 Gray (Gy) of 3-dimensional conformal RT (70 Gy) plus HT for 6 months in patients with locally advanced prostate cancer, with RT “sandwiched” by HT.8 The addition of HT for 6 months in that trial improved the 5-year overall survival rate by 10% compared with RT alone.
Neoadjuvant HT followed by RP has been viewed as a means of increasing cure rates. All 7 published randomized, placebo-controlled trials that included >1400 patients and compared HT followed by RP with RP alone reported a decrease in rates of positive surgical margins, but none demonstrated any improvement of in biochemical relapse free survival (BRFS), DFS, or overall survival.9 There also has been interest in the use of aggressive chemohormonal protocols in the neoadjuvant setting. The University of Texas M.D. Anderson Cancer Center group employed a regimen of ketoconazole, doxorubicin, vinblastine, and estramustine along with medical castration for 12 weeks preoperatively.10 Hussain et al. used combined docetaxel and estramustine for 3 to 6 cycles,11 whereas others have used docetaxel alone from 6 weeks to 6 months.12, 13 All of those Phase II trials treated small numbers of patients, and none showed a meaningful rate of pT0 disease at prostatectomy. The experience with neoadjuvant chemotherapy in breast and urothelial cancers has been that only patients with complete pathologic eradication (pT0) at the time of surgery derive a long-term clinical benefit. The use of neoadjuvant chemotherapy in locally advanced prostate cancer remains investigational. In fact, the Southwest Oncology Group (SWOG) is planning a randomized Phase III trial of neoadjuvant chemotherapy followed by RP versus RP alone in patients with high-risk, locally advanced prostate cancer.
Postoperative adjuvant RT is recommended for some men who undergo RP and are identified with ≥pT3 disease and/or macroscopically positive surgical margins. Most of those data were derived from small uncontrolled trials. Only 2 prospective randomized trials have been presented to date. EORTC 22911 randomized 1005 men with pT3N0 prostate cancer to adjuvant RT (60 Gy) or observation. At a median follow-up of 5 years, adjuvant RT decreased BRFS by 21% and DFS by 8%.14 No difference in overall survival has been observed to date. The criticism of the study is that the median PSA at time of initiation of RT was 0.2 ng/mL, which meant that a great percentage of registered patients had micrometastatic disease. The other study was reported to the American Society of Clinical Oncology in 2005 by Wiegel et al.15 The ARO 96-02/AUO AP 09/95 German collaborative group randomized 307 men with pT3 disease and postoperative undetectable PSA levels to adjuvant RT (60 Gy) versus observation. The intent-to-treat analysis showed a trend toward improved BRFS with RT. When patients were analyzed according to the actual treatment received, there was a statistically significant 21% BFFS superiority in favor of adjuvant RT. However, the large number of patients (20%) who violated their randomization limits the interpretation of those results. In view of the lack of a clear survival benefit from adjuvant RT, such practice remains investigational. Many clinicians prefer to offer salvage RT rather than adjuvant RT for select patients who develop biochemical recurrences after RP.
What little data exist relating to radical surgery for patients with cT4 prostate cancer corroborate our current findings. The 5-year results from a prospective Phase II feasibility trial (SWOG 9109) were published in 2002.16 That trial enrolled 61 eligible men with T3/T4 disease, and 55 of those 61 men underwent surgery. Four months of total androgen blockade using subcutaneous goserelin (3.6 mg monthly) and oral flutamide (250 mg daily) were delivered. With a median follow-up of 6.1 years, the 5-year overall survival rate reportedly was 90%. Although this corresponds favorably to the SEER data for combined RP and HT (82%), only 3% of the men on SWOG 9109 had cT4 disease.
In conclusion, the SEER data reveal that patients who undergo RP for cT4 prostate cancer may have increased survival compared with patients who receive RT alone or HT alone, and their survival is comparable to that achieved by patients who receive combined RT and HT. However the benefit of RP appears to be limited to a relatively small subset of patients who have regional lymph node extension. Adjuvant therapy does not appear to improve survival for patients after RP for cT4 prostate cancer.
- 1Standards for Cancer Registries, volume II: Data Standards and Data dictionary. 10th ed. Springfield, IL: North American Association of Central Cancer Registries; 2004., .
- 6Updated results of the Phase III Radiation Therapy Oncology Group (RTOG) trial 85-31 evaluating the potential benefit of androgen suppression following standard radiation therapy for unfavorable prognosis carcinoma of the prostate. Int J Radiat Oncol Biol Phys. 2001; 49: 937–946., , , et al.
- 15Phase III results of adjuvant radiotherapy versus “wait and see” in patients with pT3 prostate cancer following radical prostatectomy (ARO 96-02/AUO AP 09/95). Proc Am Soc Clin Oncol. 2005; 23: 381s. Abstract 4513., , , et al.