- Email: [email protected]
Redefining Clinically Significant Castration Levels in Patients With Prostate Cancer Receiving Continuous Androgen Deprivation Therapy
Redefining Clinically Significant Castration Levels in Patients With Prostate Cancer Receiving Continuous Androgen Deprivation Therapy Juan Morote, Anna Orsola,* Jacques Planas, Enrique Trilla, Carles X. Raventós, Lluís Cecchini and Roberto Catalán From the Department of Urology, Vall d’Hebron Hospital and Autonoma University of Barcelona School of Medicine, Barcelona, Spain
Purpose: We determined the testosterone castration level with clinical relevance in patients with prostate cancer on continuous androgen deprivation therapy. Secondary objectives were to analyze the role of associated bicalutamide in breakthrough increases of serum testosterone in these patients and the possible benefit of maximal androgen blockade. Materials and Methods: Serum testosterone was determined 3 times (in 6 months) in 73 patients with nonmetastatic prostate cancer treated with medical castration, 28 (38.4%) of whom also received bicalutamide (maximal androgen blockade). During a mean followup of 51 months (range 12 to 240) 41 (67.1%) events of androgen independent progression were identified, and correlated with breakthrough testosterone increases of 50 ng/dl (classic level) and 20 ng/dl (surgical castration level). Results: Testosterone was less than 20 ng/dl in all determinations in 32 patients (43.6%). Breakthrough increases between 20 and 50 ng/dl were observed in 23 patients (31.5%), and increases greater than 50 ng/dl were observed in the remaining 18 (24.7%). The lowest testosterone level with a significant impact on survival free of androgen independent progression was 32 ng/dl. Mean survival free of androgen independent progression in patients with breakthrough increases greater than 32 ng/dl was 88 months (95% CI 55–121) while it was 137 months (95% CI 104 –170) in those without breakthrough increases (p ⬍0.03). Patients on maximal androgen blockade had an incidence of testosterone increase similar to those receiving monotherapy. However, maximal androgen blockade provided a significantly longer survival free of androgen independent progression in those with breakthrough increases greater than 50 ng/dl. Conclusions: In the current report the lowest testosterone castration level with clinical relevance in medically castrated patients with prostate cancer was 32 ng/dl. Breakthrough increases greater than this threshold predicted a lower survival free of androgen independent progression. Maximal androgen blockade might benefit medically castrated cases of prostate cancer with breakthrough increases of more than 50 ng/dl. Key Words: testosterone, prostatic neoplasms, androgen antagonists, castration
ndrogen withdrawal has been the standard systemic therapy for advanced prostate cancer since the demonstration of hormone dependence of prostate cancer by Huggins and Hodges.1 During the last 20 years surgical castration has been replaced by the medical use of LH-RH agonists as an effective and well tolerated treatment option for PCa.2 After the characterization of the structure of luteinizing hormone releasing hormone by Schally et al,3 several synthetic agonist of LH-RH have been developed and tested for clinical use and found to be as effective as bilateral orchiectomy in treating PCa.4 In addition, these agents avoid the increase in thromboembolic events associated with estrogens.5 The goal of surgical or chemical castration is to achieve low testosterone levels. The importance of low testosterone levels is based on the knowledge that ineffective androgen suppression therapy results in higher PCa mortality.
A
Submitted for publication January 22, 2007. * Correspondence and requests for reprints: Department of Urology, Vall d=Hebron Hospital, Po Vall d=Hebron 119-129, Barcelona 08035, Spain (e-mail: [email protected]; [email protected]).
See Editorial on page 1148. For another article on a related topic see page 1521.
0022-5347/07/1784-1290/0 THE JOURNAL OF UROLOGY® Copyright © 2007 by AMERICAN UROLOGICAL ASSOCIATION
LH-RH agonist response in patients with PCa has conventionally been considered achieved when testosterone levels are less than 50 ng/dl. The availability of new measuring techniques and of new generations of LH-RH agonists has challenged this definition. Castration levels have been redefined as appropriate when reaching 20 ng/dl based on levels obtained by surgical castration.6 However, neither therapeutic effect nor clinical benefit nor survival advantage of this new threshold has been demonstrated to date. In consequence, controversy exists regarding the appropriate castration level of testosterone. Sharifi and Browneller postulated that serum testosterone is suppressed if there are 2 consecutive determinations on separate days less than 50 ng/dl.7 This definition has been applied in many clinical trials and most physicians consider LH-RH agonists efficacious when achieving this level. However, a testosterone castration cutoff of 50 ng/dl is derived from the use of old assay methods.8 In 1996 a fully automated immunoassay analyzer using chemiluminescent technology was approved for clinical use.9 This technique allows more accurate serum testosterone measurements and currently the sensitivity of this assay is as accurate as 0.1 ng/dl.10 Oefelein et al conducted a study to reevaluate the castration levels achieved
1290
Vol. 178, 1290-1295, October 2007 Printed in U.S.A. DOI:10.1016/j.juro.2007.05.129
BREAKTHROUGH INCREASES OF TESTOSTERONE IN PROSTATE CANCER with bilateral orchiectomy using a chemiluminescent immunoassay method.6 Median serum testosterone in a group of 35 patients with advanced prostate cancer who underwent bilateral orchiectomy was 15 ng/dl (95% CI between 12 and 17 ng/dl) with minimum and maximum values of 10 and 30 ng/dl, respectively. These authors postulated that the lower the better when trying to achieve castration levels of testosterone, and suggested that 20 ng/dl is the most appropriate cut point to define castration. Although most patients reach testosterone castration levels after 3 to 4 weeks on LH-RH agonists, there is always the possibility of hormonal escape during LH-RH agonist therapy. A hormonal escape is defined as an increase in serum testosterone after castration levels have been achieved. Although it seems an important clinical consideration, this event is not recognized if serum testosterone is not systematically monitored. The frequency of breakthrough responses and patients with an acute-on-chronic response has been recently analyzed by Tombal.11 These hormonal escapes are indicators of ineffective testosterone suppression but are often unrecognized by the physician. However, there are no data analyzing the impact of serum testosterone levels and breakthrough increases on prostate cancer clinical outcome. In the present study we analyze the rate of breakthrough increases in serum testosterone in patients receiving 3 months of depot LH-RH agonist according to the proposed castration level of 20 ng/dl and the classic level of 50 ng/dl. The primary objective was to determine the lowest serum testosterone castration level with clinical relevance, the end point being survival free of AIP. The secondary objective was to analyze the relationship between bicalutamide associated treatment to LH-RH agonist and breakthrough increases of serum testosterone, and the possible benefit of mAb in these patients. MATERIALS AND METHODS Patient Selection, Treatment and Followup Since October 2001 in our department, serum testosterone is determined with PSA in patients on ADT. A group of 73 patients with a histological diagnosis of prostate cancer treated with 3 months of depot LH-RH agonist were selected from our database. Those with more than 1 year of followup and at least 3 serum testosterone determinations were included in the study. Patient characteristics are summarized in table 1. All patients had nonmetastatic disease, and stag-
TABLE 1. Patient characteristics No. pts Mean pt age ⫾ SD (range) No. clinical stage at diagnosis (%): T1–2 N0 M0 T3–4 N0 M0 No. biopsy Gleason score (%): 2–7 8–10 Mean PSA at diagnosis ⫾ SD (range) No. initial treatment (%): Radical prostatectomy ADT No. ADT modality (%): LH-RH agonist alone LH-RH agonist ⫹ bicalutamide Mean mos on ADT ⫾ SD (range)
73 70.7 ⫾ 7.6 63 10
(54–88) (86.3) (13.7)
50 (68.5) 23 (31.5) 81.2 ⫾ 184.9 (4.3–1.320) 23 50 45 28 54.1 ⫾ 43.8
(31.5) (68.5) (61.6) (38.4) (12–238)
1291
ing at diagnosis included bone scintigraphy if serum PSA was higher than 10 ng/ml and abdominal computerized tomography if serum PSA was higher than 20 ng/ml. In 50 patients (68.5%) ADT was the initial treatment for prostate cancer. In 23 patients (31.5%) the primary treatment for prostate cancer was radical prostatectomy. In this group ADT was started in an adjuvant manner because of PSA progression in 20 patients, seminal vesicle invasion in 2 and positive lymph nodes in 1 patient. All patients received 50 mg bicalutamide per day 2 weeks before the first LH-RH agonist administration. Patients were instructed to administer a new dose of the 3-month depot LH-RH agonist every 90 days. A subset of 28 patients continued treatment with bicalutamide in a maximal androgen blockade manner. The decision to maintain or withdraw the antiandrogen was made based on side effects and after informed consent. No patient in this series received other modalities of treatment for prostate cancer. Mean followup after ADT was 54.1 months (range 13 to 240) and during this period 41 events of AIP (67.1%) were detected. Testosterone Serum Determinations Serum determinations of testosterone were performed every 6 months at the same time as routine PSA control. Blood samples were obtained between weeks 8 and 12 after the 3-month depot LH-RH agonist injection, between 8:00 and 10:00 am. A chemiluminescent assay, the Immulite® 2500, was used to determine serum testosterone, the lower serum testosterone informed value being less than 15 ng/dl. Normal levels defined by the laboratory were between 212 and 742 ng/dl in men, and between 49 and 102 ng/dl in women. Definition of Breakthrough Increases of Serum Testosterone Any increase in serum testosterone greater than 20 ng/ml was considered a breakthrough response. Patients were classified according to cutoff points of testosterone in 3 groups, with group 1—patients with all 3 serum testosterone determinations lower than 20 ng/dl, group 2—patients with breakthrough increases of serum testosterone between 20 and 50 ng/dl, and group 3—patients with breakthrough increases greater than 50 ng/dl. Definition of Androgen Independent Progression AIP was defined as 3 consecutive PSA increases after the nadir. When a first increase in PSA was detected 2 more PSA determinations were performed in the following 2 months. Survival free of AIP was established as the period between the day of the first LH-RH agonist injection and the day of the first PSA increase after the nadir PSA. Statistical Analysis Mean or median, standard deviation and range were used to define quantitative variables, and rates were used to define qualitative ones. The chi-square test and the nonparametric Mann-Whitney U test were used to relate qualitative and quantitative variables, respectively. Kaplan-Meier curves and the log rank test were used to analyze the cumulative survival of free PSA progression. Recoding procedures were made in quantitative variables to generate categorical ones to best characterize multivariate Cox regression analysis
BREAKTHROUGH INCREASES OF TESTOSTERONE IN PROSTATE CANCER
using the forward stepwise conditional method. The variables included in the analysis were age 70 years or younger vs more than 70 years, clinical stage T1–2 vs T3– 4, Gleason score 2–7 vs 8 –10, androgen deprivation therapy in terms of LH-RH agonist alone vs maximal androgen blockade (maximal androgen blockade as LH-RH plus bicalutamide), PSA at diagnosis 20 or less vs more than 20 ng/ml, breakthrough increases of testosterone 20 or less vs more than 20 ng/dl, and breakthrough increases of testosterone 50 or less vs more than 50 ng/dl. The lowest serum testosterone threshold with a significant impact on survival free of AIP was estimated analyzing the log rank test for every threshold in decreasing order. SPSS® V.12 software was used for all analysis. RESULTS Distribution of Patients According to Serum Testosterone Behavior In 33 of the 73 patients (48.6%) all 3 serum determinations of testosterone remained less than 20 ng/dl. In 23 patients (31.5%) at least 1 serum determination of testosterone was between 20 and 50 ng/dl. In 18 patients (24.7%) at least 1 serum determination of testosterone was greater than 50 ng/dl. In the remaining 3 patients (4.1%) all 3 serum determinations of testosterone were greater than 50 ng/dl, ranging between 54.5 and 115.4 ng/dl. Factors Related to Survival Free of AIP Univariate analysis with the log rank test was done using categorical variables as shown in table 2. Gleason score (p ⫽ 0.0096) and breakthrough increases of serum testosterone greater than 50 ng/dl (p ⫽ 0.0054) were the only 2 variables related with the survival free of AIP using the listed variables. In contrast, if the threshold of serum testosterone was established at 20 ng/dl no significant relationship between breakthrough increases and survival free of AIP was observed. Multivariate Cox regression analysis using the forward stepwise conditional method showed that a breakthrough increase in serum testosterone greater than 50 ng/dl was, from this group of variables, the only predictive factor of survival free of AIP, p ⫽ 0.008, with a hazard ratio of 2.8 (95% CI 1.3–5.9). Threshold of Serum Testosterone Breakthrough Increases With Clinical Impact The Kaplan-Meier curves analyzing survival free of AIP in the established groups according to the behavior of serum testosterone is presented in figure 1. Mean survival free of AIP was 106 months (95% CI 97–115) in group 1 (all 3 serum
TABLE 2. Univariate analysis relating dichotomic variables included in the study and survival free of AIP
Age (70 or younger vs older than 70 yrs) Initial clinical stage (T1–2 vs T3–4) Biopsy Gleason score (2–6 vs 7–10) LH-RH agonist (⫾bicalutamide) Initial PSA (20 or less vs greater than 20 ng/ml) Testosterone breakthrough increases greater than 20 ng/dl Testosterone breakthrough increases greater than 50 ng/dl
Log Rank
p Value
3.27 0.01 6.71 0.99 0.10 1.78
0.0723 0.9095 0.0096 0.3188 0.7483 0.1819
7.74
0.0054
1,0
Cum survival free of PSA progression
1292
p 0.0207
0,8
Group 1
0,6
Group 2
0,4
0,2
Group 3
0,0 0
24
48
72
96
120
144
168
192 216
240
Months under ADT FIG. 1. Survival free of AIP according to serum testosterone behavior. Group 1, patients with all 3 serum testosterone determinations less than 20 ng/dl. Group 2, patients with breakthrough increases between 20 and 50 ng/dl. Group 3, patients with breakthrough increases greater than 50 ng/dl.
determinations less than 20 ng/dl), 90 months (95% CI 73–108) in group 2 (any increase between 20 and 50 ng/dl) and 72 months (95% CI 33–111) in group 3 (any increase greater than 50 ng/dl). Because patients with all 3 determinations less than 20 ng/dl (group 1) had a longer survival free of AIP than the other 2 groups, the clinically significant threshold of testosterone was investigated. The lowest serum testosterone threshold that was able to significantly distinguish groups related with the survival free of AIP was 32 ng/dl, p ⫽ 0.0258 (fig. 2). Patients with all 3 determinations of serum testosterone less than 32 ng/dl had a mean survival free of AIP of 137 (95% CI 104 –170) vs 88 months (95% CI 55–121) for those with any breakthrough increase greater than 32 ng/dl, p ⬍0.03. Relationship Between Breakthrough Increases and ADT Modality A cross tabulation analysis between ADT modality and breakthrough increases of serum testosterone greater than any threshold (20, 32 or 50 ng/dl) revealed no significant relation between treatment with bicalutamide and the incidence of breakthrough increases in serum testosterone. In the subset of patients on maximal androgen blockade the rate of breakthrough increases of serum testosterone greater than 50 ng/dl was 75.6% (33 of 45), similar to the 75.0% (21 of 28) in those patients treated with medical castration, p ⫽ 0.957. The clinical impact of ADT modality in those in whom breakthrough increases of serum testosterone developed at some point was analyzed. Taking the conventional threshold of 50 ng/dl, mean survival free of PSA progression was 115 months (95% CI 53–172) when receiving maximal androgen blockade as opposed to 32 months (95% CI 19 – 45) when receiving LH-RH agonist alone (p ⫽ 0.0249, fig. 3, A).
BREAKTHROUGH INCREASES OF TESTOSTERONE IN PROSTATE CANCER
Cumulate survival free of AIP (%)
1,0
p = 0.0258
0,8
1293
with medical castration has been established. In recent years the availability of new and more accurate assay methods has enabled the detection of serum testosterone levels below the threshold of 20 ng/dl,9 which is much lower than the conventional 50 ng/dl castration level defined after the phase III study for the Food and Drug Administration licensing of 22.5 mg leuprolide depot.13 Using this more strin-
Group 1
0,6
0,4
Group 2
0,2
0,0 0
50
100
150
200
250
Follow up (months) FIG. 2. Survival free of AIP analysis according to serum testosterone behavior in relation to lowest castrate threshold established at 32 ng/dl. Group 1, patients with all 3 serum testosterone determinations less than 32 ng/dl. Group 2, patients with breakthrough increases greater than 32 ng/dl.
Taking the 32 ng/dl threshold, mean survival free of AIP for patients on mAb was 98 months (95% CI 57–138) while it was 63 months (95% CI 39 – 87) for patients receiving LH-RH agonist alone, which did not reach significance (p ⫽ 0.3350). In patients without breakthrough increases greater than 32 ng/dl mean survival was 126 (95% CI 86 –167) (mAb) and 104 months (95% CI 77–132) (LH-RH agonist alone) (p ⫽ 0.2801), respectively. DISCUSSION This study demonstrates that breakthrough increases of serum testosterone in medically castrated patients with prostate cancer are not only frequent but also have clinical implications regarding PSA progression. In addition, the results suggest that 32 ng/dl might well be the lowest serum testosterone threshold with a clinical impact in these patients. Patients with breakthrough increases in serum testosterone greater than 32 ng/dl had a significantly shorter survival free of AIP (88 months) than those patients without these events (137 months). Finally, observations in the report imply that, at least for the highest and classic range of testosterone of 50 ng/dl, maximal androgen blockade (by adding bicalutamide to LH-RH agonist) might have a beneficial role. To our knowledge this is the first report to establish a direct relationship between testosterone increases and androgen independent progression. The importance of effective testosterone suppression is frequently emphasized on the basis that ineffective androgen suppression may result in increased prostate cancer mortality. However, this observation is based on more than 30-year-old studies with diethylstilbestrol.12 Currently no evidence of the clinical implications of breakthrough increases of testosterone when treated
FIG. 3. Survival free of AIP according to ADT modality and serum testosterone threshold established at 50 ng/dl. Group 1, patients treated with LH-RH agonist plus bicalutamide. Group 2, patients treated with LH-RH agonist alone. A, patients with breakthrough increases of serum testosterone greater than 50 ng/dl. B, patients without breakthrough increases.
1294
BREAKTHROUGH INCREASES OF TESTOSTERONE IN PROSTATE CANCER
gent definition of 20 ng/dl, recent publications have shown that several patients treated with current LH-RH agonist formulations do not achieve this castration level, and show breakthroughs and acute-on-chronic responses.11 In addition, up to 13% of men on depot LH-RH agonists will have testosterone values between 20 and 50 ng/dl.14 This has triggered the question of whether achieving testosterone levels less than 50 ng/dl corresponds to optimal castration, and if 20 ng/dl, based on the level obtained with orchiectomy as a benchmark, would be a better threshold for defining an appropriate castration and treatment response.6 The present study analyzes 3 consecutive serum testosterone determinations performed at 6-month intervals in medically castrated patients with prostate cancer. The rate of patients with all 3 serum testosterone determinations greater than 20 ng/dl was 43.8%, whereas breakthrough increases of serum testosterone between 20 and 50 ng/dl were detected in 31.5% of patients, and greater than 50 ng/dl in 24.7%. In a recent review Tombal reported a rate of breakthrough increase in serum testosterone greater than 20 ng/dl in 13% to 35% of patients treated with LH-RH agonists, and greater than 50 ng/dl in 2% to 13% of these patients.11 Of note, whether this discrepancy in hormonal levels translates into a therapeutic effect and a survival advantage or disadvantage is unknown. In fact, clinical trials comparing outcomes between the various forms of hormonal therapy have reported no survival advantage for orchiectomy that achieves lower testosterone levels than monthly LH-RH agonist (3.6 mg goserelin acetate).4,15 However, this must be interpreted cautiously because of a relatively short median followup (less than 15 months), and because neither the British nor the United States prostate study groups were significantly powered to detect small differences in outcome.4,15 In the present report we failed to demonstrate a direct clinical benefit of the lower cutoff value of 20 ng/dl. However, 32 ng/dl, which is considerably lower than 50 ng/dl, showed a significant impact on survival free of AIP. Of note, in classic report by Oefelein et al in which the idea of the lower the better was postulated, values of up to 30 ng/dl were also observed in surgically castrated cases.6 Central to the debate regarding therapeutic castration is the potential beneficial role of maximal androgen blockade compared to castration alone. A modest overall and cancer specific survival has been shown on meta-analysis but associated with increased adverse events and reduced quality of life.16 The National Comprehensive Cancer Network recommends that if the LH-RH agonist therapy testosterone reaches castration values less than 20 ng/dl, not to add an antiandrogen. However, if the serum testosterone level is greater than 20 ng/dl and the patient refuses surgical castration, the addition of an antiandrogen should be considered.17 These guidelines are recommendations but they have not been clinically confirmed.15 Observations in our report support a potential protective effect of the addition of bicalutamide. Time to AIP was longer in patients on mAb compared to those treated with LH-RH analogues alone, which was the case for those with (98 vs 63 months) or without (126 vs 104 months) breakthrough increases in testosterone greater than 32 ng/dl. However, this finding only reached statistical significance when using the classic 50 ng/dl threshold (115 vs 32 months, p ⫽ 0.0249).
Collectively the findings of the current report prove that testosterone breakthrough increase has prognostic value in medically castrated patients based on the direct evidence of longer AIP. Our findings establish for the first time a cutoff value, 32 ng/dl, with significant clinical impact because increases greater than this threshold were related to a lower survival free of AIP. Our data suggest a rationale for the use of maximal androgen blockade with bicalutamide when breakthrough increases of serum testosterone greater than 50 ng/dl are detected, which is in accordance with the National Comprehensive Cancer Network guidelines and based on possible clinical benefit in these patients. CONCLUSIONS Because physicians today are not routinely evaluating the testosterone level when initiating and following ADT, we suggest that routine measurement of serum testosterone should become part of clinical practice when evaluating the effects of hormonal therapy. A reasonable option to detect these breakthrough increases would be to monitor testosterone levels at PSA determination. Apart from having prognostic implications, knowing the level of testosterone would allow redosing the LH-RH agonist accordingly and considering maximal androgen blockade. A prospective, randomized and carefully designed trial contemplating clinical progression and specific mortality as the primary end point would be required to confirm these findings and reassess the cutoff level.
Abbreviations and Acronyms ADT AIP LH-RH mAb PCa PSA
⫽ ⫽ ⫽ ⫽ ⫽ ⫽
androgen deprivation therapy androgen independent progression luteinizing hormone releasing hormone maximal androgen blockade prostate cancer prostate specific antigen
REFERENCES 1.
2.
3.
4.
5. 6.
7.
Huggins C and Hodges C: Studies on prostate cancer: the effect of castration, estrogen and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate. Cancer Res 1941; 1: 293. Mongiat-Artus P and Teillac P: Role of luteinising hormone releasing hormone (LHRH) agonist in the management of prostate cancer. Eur Urol, suppl, 2005; 4: 4. Schally AV, Kastin AJ and Arimura A: Hypothalamic folliclestimulating hormone (FSH) and luteinizing hormone (LH)regulating hormone: structure, physiology and clinical studies. Fertil Steril 1971; 22: 703. Kaisary AV, Tyrrell CJ, Peeling WB and Griffiths K: Comparison of LHRH analogue (Zoladex) with orchiectomy in patients with metastatic prostatic carcinoma. Br J Urol 1991; 67: 502. The Leuprolide Study Group: Leuprolide versus diethylstilbestrol for metastatic prostate cancer. N Engl J Med 1984; 311: 1281. Oefelein MG, Feng A, Scolieri MJ, Ricchiutti D and Resnick MI: Reassessment of the definition of castrate levels of testosterone: implications for clinical decision making. Urology 2000; 56: 1021. Sharifi R and Browneller R: Serum testosterone suppression and potential for agonistic stimulation during chronic treat-
BREAKTHROUGH INCREASES OF TESTOSTERONE IN PROSTATE CANCER ment with monthly and 3-month depot formulations of leuprolide acetate for advanced prostate cancer. J Urol 2002; 168: 1001. 8. Wilke TJ and Utley DJ: Total testosterone, free-androgen index, calculated free testosterone, and free testosterone by analog RIA compared in hirsute women and otherwisenormal women with altered binding of sex-hormonebinding globulin. Clin Chem 1987; 33: 1372. 9. Wheeler MJ, D’Souza A, Matadeen J and Croos P: Ciba Corning ACS:180 testosterone assay evaluated. Clin Chem 1996; 42: 1445. 10. Zherdev AV, Byzova NA, Izumrudov VA and Dzantiev BB: Rapid polyelectrolyte-based immunofiltration technique for testosterone detection in serum samples. Analyst 2003; 128: 1275. 11. Tombal B: Appropriate castration with luteinising hormone releasing hormone (LHRH) agonist: what is the optimal level of testosterone. Eur Urol, suppl, 2005; 4: 14. 12. Jordan WP Jr, Blackard CE and Byar DP: Reconsideration of orchiectomy in the treatment of advanced prostatic carcinoma. South Med J 1977; 70: 1411.
13.
1295
Sharifi R, Bruskewitz RC, Gittleman MC, Graham SD Jr, Hudson PB and Stein B: Leuprorelin acetate 22.5 mg 12week depot formulation in the treatment of patients with advanced prostate cancer. Clin Ther 1996; 18: 647. 14. Oefelein MG and Cornum R: Failure to achieve castrate levels of testosterone during luteinizing hormone releasing hormone agonist therapy: the case for monitoring serum testosterone and a treatment decision algorithm. J Urol 2000; 164: 726. 15. Vogelzang NJ, Chodak GW, Soloway MS, Block NL, Schellhammer PF, Smith JA Jr et al: Goserelin versus orchiectomy in the treatment of advanced prostate cancer: final results of a randomized trial. Zoladex Prostate Study Group. Urology 1995; 46: 220. 16. Schmitt B, Bennett C, Seidenfeld J, Samson D and Wilt T: Maximal androgen blockade for advanced prostate cancer. Cochrane Database Syst Rev 2000; 2: CD001526. 17. Millikan R and Logothetis C: Update of the NCCN guidelines for treatment of prostate cancer. Oncology (Williston Park) 1997; 11: 180.
Purpose: We determined the testosterone castration level with clinical relevance in patients with prostate cancer on continuous androgen deprivation therapy. Secondary objectives were to analyze the role of associated bicalutamide in breakthrough increases of serum testosterone in these patients and the possible benefit of maximal androgen blockade. Materials and Methods: Serum testosterone was determined 3 times (in 6 months) in 73 patients with nonmetastatic prostate cancer treated with medical castration, 28 (38.4%) of whom also received bicalutamide (maximal androgen blockade). During a mean followup of 51 months (range 12 to 240) 41 (67.1%) events of androgen independent progression were identified, and correlated with breakthrough testosterone increases of 50 ng/dl (classic level) and 20 ng/dl (surgical castration level). Results: Testosterone was less than 20 ng/dl in all determinations in 32 patients (43.6%). Breakthrough increases between 20 and 50 ng/dl were observed in 23 patients (31.5%), and increases greater than 50 ng/dl were observed in the remaining 18 (24.7%). The lowest testosterone level with a significant impact on survival free of androgen independent progression was 32 ng/dl. Mean survival free of androgen independent progression in patients with breakthrough increases greater than 32 ng/dl was 88 months (95% CI 55–121) while it was 137 months (95% CI 104 –170) in those without breakthrough increases (p ⬍0.03). Patients on maximal androgen blockade had an incidence of testosterone increase similar to those receiving monotherapy. However, maximal androgen blockade provided a significantly longer survival free of androgen independent progression in those with breakthrough increases greater than 50 ng/dl. Conclusions: In the current report the lowest testosterone castration level with clinical relevance in medically castrated patients with prostate cancer was 32 ng/dl. Breakthrough increases greater than this threshold predicted a lower survival free of androgen independent progression. Maximal androgen blockade might benefit medically castrated cases of prostate cancer with breakthrough increases of more than 50 ng/dl. Key Words: testosterone, prostatic neoplasms, androgen antagonists, castration
ndrogen withdrawal has been the standard systemic therapy for advanced prostate cancer since the demonstration of hormone dependence of prostate cancer by Huggins and Hodges.1 During the last 20 years surgical castration has been replaced by the medical use of LH-RH agonists as an effective and well tolerated treatment option for PCa.2 After the characterization of the structure of luteinizing hormone releasing hormone by Schally et al,3 several synthetic agonist of LH-RH have been developed and tested for clinical use and found to be as effective as bilateral orchiectomy in treating PCa.4 In addition, these agents avoid the increase in thromboembolic events associated with estrogens.5 The goal of surgical or chemical castration is to achieve low testosterone levels. The importance of low testosterone levels is based on the knowledge that ineffective androgen suppression therapy results in higher PCa mortality.
A
Submitted for publication January 22, 2007. * Correspondence and requests for reprints: Department of Urology, Vall d=Hebron Hospital, Po Vall d=Hebron 119-129, Barcelona 08035, Spain (e-mail: [email protected]; [email protected]).
See Editorial on page 1148. For another article on a related topic see page 1521.
0022-5347/07/1784-1290/0 THE JOURNAL OF UROLOGY® Copyright © 2007 by AMERICAN UROLOGICAL ASSOCIATION
LH-RH agonist response in patients with PCa has conventionally been considered achieved when testosterone levels are less than 50 ng/dl. The availability of new measuring techniques and of new generations of LH-RH agonists has challenged this definition. Castration levels have been redefined as appropriate when reaching 20 ng/dl based on levels obtained by surgical castration.6 However, neither therapeutic effect nor clinical benefit nor survival advantage of this new threshold has been demonstrated to date. In consequence, controversy exists regarding the appropriate castration level of testosterone. Sharifi and Browneller postulated that serum testosterone is suppressed if there are 2 consecutive determinations on separate days less than 50 ng/dl.7 This definition has been applied in many clinical trials and most physicians consider LH-RH agonists efficacious when achieving this level. However, a testosterone castration cutoff of 50 ng/dl is derived from the use of old assay methods.8 In 1996 a fully automated immunoassay analyzer using chemiluminescent technology was approved for clinical use.9 This technique allows more accurate serum testosterone measurements and currently the sensitivity of this assay is as accurate as 0.1 ng/dl.10 Oefelein et al conducted a study to reevaluate the castration levels achieved
1290
Vol. 178, 1290-1295, October 2007 Printed in U.S.A. DOI:10.1016/j.juro.2007.05.129
BREAKTHROUGH INCREASES OF TESTOSTERONE IN PROSTATE CANCER with bilateral orchiectomy using a chemiluminescent immunoassay method.6 Median serum testosterone in a group of 35 patients with advanced prostate cancer who underwent bilateral orchiectomy was 15 ng/dl (95% CI between 12 and 17 ng/dl) with minimum and maximum values of 10 and 30 ng/dl, respectively. These authors postulated that the lower the better when trying to achieve castration levels of testosterone, and suggested that 20 ng/dl is the most appropriate cut point to define castration. Although most patients reach testosterone castration levels after 3 to 4 weeks on LH-RH agonists, there is always the possibility of hormonal escape during LH-RH agonist therapy. A hormonal escape is defined as an increase in serum testosterone after castration levels have been achieved. Although it seems an important clinical consideration, this event is not recognized if serum testosterone is not systematically monitored. The frequency of breakthrough responses and patients with an acute-on-chronic response has been recently analyzed by Tombal.11 These hormonal escapes are indicators of ineffective testosterone suppression but are often unrecognized by the physician. However, there are no data analyzing the impact of serum testosterone levels and breakthrough increases on prostate cancer clinical outcome. In the present study we analyze the rate of breakthrough increases in serum testosterone in patients receiving 3 months of depot LH-RH agonist according to the proposed castration level of 20 ng/dl and the classic level of 50 ng/dl. The primary objective was to determine the lowest serum testosterone castration level with clinical relevance, the end point being survival free of AIP. The secondary objective was to analyze the relationship between bicalutamide associated treatment to LH-RH agonist and breakthrough increases of serum testosterone, and the possible benefit of mAb in these patients. MATERIALS AND METHODS Patient Selection, Treatment and Followup Since October 2001 in our department, serum testosterone is determined with PSA in patients on ADT. A group of 73 patients with a histological diagnosis of prostate cancer treated with 3 months of depot LH-RH agonist were selected from our database. Those with more than 1 year of followup and at least 3 serum testosterone determinations were included in the study. Patient characteristics are summarized in table 1. All patients had nonmetastatic disease, and stag-
TABLE 1. Patient characteristics No. pts Mean pt age ⫾ SD (range) No. clinical stage at diagnosis (%): T1–2 N0 M0 T3–4 N0 M0 No. biopsy Gleason score (%): 2–7 8–10 Mean PSA at diagnosis ⫾ SD (range) No. initial treatment (%): Radical prostatectomy ADT No. ADT modality (%): LH-RH agonist alone LH-RH agonist ⫹ bicalutamide Mean mos on ADT ⫾ SD (range)
73 70.7 ⫾ 7.6 63 10
(54–88) (86.3) (13.7)
50 (68.5) 23 (31.5) 81.2 ⫾ 184.9 (4.3–1.320) 23 50 45 28 54.1 ⫾ 43.8
(31.5) (68.5) (61.6) (38.4) (12–238)
1291
ing at diagnosis included bone scintigraphy if serum PSA was higher than 10 ng/ml and abdominal computerized tomography if serum PSA was higher than 20 ng/ml. In 50 patients (68.5%) ADT was the initial treatment for prostate cancer. In 23 patients (31.5%) the primary treatment for prostate cancer was radical prostatectomy. In this group ADT was started in an adjuvant manner because of PSA progression in 20 patients, seminal vesicle invasion in 2 and positive lymph nodes in 1 patient. All patients received 50 mg bicalutamide per day 2 weeks before the first LH-RH agonist administration. Patients were instructed to administer a new dose of the 3-month depot LH-RH agonist every 90 days. A subset of 28 patients continued treatment with bicalutamide in a maximal androgen blockade manner. The decision to maintain or withdraw the antiandrogen was made based on side effects and after informed consent. No patient in this series received other modalities of treatment for prostate cancer. Mean followup after ADT was 54.1 months (range 13 to 240) and during this period 41 events of AIP (67.1%) were detected. Testosterone Serum Determinations Serum determinations of testosterone were performed every 6 months at the same time as routine PSA control. Blood samples were obtained between weeks 8 and 12 after the 3-month depot LH-RH agonist injection, between 8:00 and 10:00 am. A chemiluminescent assay, the Immulite® 2500, was used to determine serum testosterone, the lower serum testosterone informed value being less than 15 ng/dl. Normal levels defined by the laboratory were between 212 and 742 ng/dl in men, and between 49 and 102 ng/dl in women. Definition of Breakthrough Increases of Serum Testosterone Any increase in serum testosterone greater than 20 ng/ml was considered a breakthrough response. Patients were classified according to cutoff points of testosterone in 3 groups, with group 1—patients with all 3 serum testosterone determinations lower than 20 ng/dl, group 2—patients with breakthrough increases of serum testosterone between 20 and 50 ng/dl, and group 3—patients with breakthrough increases greater than 50 ng/dl. Definition of Androgen Independent Progression AIP was defined as 3 consecutive PSA increases after the nadir. When a first increase in PSA was detected 2 more PSA determinations were performed in the following 2 months. Survival free of AIP was established as the period between the day of the first LH-RH agonist injection and the day of the first PSA increase after the nadir PSA. Statistical Analysis Mean or median, standard deviation and range were used to define quantitative variables, and rates were used to define qualitative ones. The chi-square test and the nonparametric Mann-Whitney U test were used to relate qualitative and quantitative variables, respectively. Kaplan-Meier curves and the log rank test were used to analyze the cumulative survival of free PSA progression. Recoding procedures were made in quantitative variables to generate categorical ones to best characterize multivariate Cox regression analysis
BREAKTHROUGH INCREASES OF TESTOSTERONE IN PROSTATE CANCER
using the forward stepwise conditional method. The variables included in the analysis were age 70 years or younger vs more than 70 years, clinical stage T1–2 vs T3– 4, Gleason score 2–7 vs 8 –10, androgen deprivation therapy in terms of LH-RH agonist alone vs maximal androgen blockade (maximal androgen blockade as LH-RH plus bicalutamide), PSA at diagnosis 20 or less vs more than 20 ng/ml, breakthrough increases of testosterone 20 or less vs more than 20 ng/dl, and breakthrough increases of testosterone 50 or less vs more than 50 ng/dl. The lowest serum testosterone threshold with a significant impact on survival free of AIP was estimated analyzing the log rank test for every threshold in decreasing order. SPSS® V.12 software was used for all analysis. RESULTS Distribution of Patients According to Serum Testosterone Behavior In 33 of the 73 patients (48.6%) all 3 serum determinations of testosterone remained less than 20 ng/dl. In 23 patients (31.5%) at least 1 serum determination of testosterone was between 20 and 50 ng/dl. In 18 patients (24.7%) at least 1 serum determination of testosterone was greater than 50 ng/dl. In the remaining 3 patients (4.1%) all 3 serum determinations of testosterone were greater than 50 ng/dl, ranging between 54.5 and 115.4 ng/dl. Factors Related to Survival Free of AIP Univariate analysis with the log rank test was done using categorical variables as shown in table 2. Gleason score (p ⫽ 0.0096) and breakthrough increases of serum testosterone greater than 50 ng/dl (p ⫽ 0.0054) were the only 2 variables related with the survival free of AIP using the listed variables. In contrast, if the threshold of serum testosterone was established at 20 ng/dl no significant relationship between breakthrough increases and survival free of AIP was observed. Multivariate Cox regression analysis using the forward stepwise conditional method showed that a breakthrough increase in serum testosterone greater than 50 ng/dl was, from this group of variables, the only predictive factor of survival free of AIP, p ⫽ 0.008, with a hazard ratio of 2.8 (95% CI 1.3–5.9). Threshold of Serum Testosterone Breakthrough Increases With Clinical Impact The Kaplan-Meier curves analyzing survival free of AIP in the established groups according to the behavior of serum testosterone is presented in figure 1. Mean survival free of AIP was 106 months (95% CI 97–115) in group 1 (all 3 serum
TABLE 2. Univariate analysis relating dichotomic variables included in the study and survival free of AIP
Age (70 or younger vs older than 70 yrs) Initial clinical stage (T1–2 vs T3–4) Biopsy Gleason score (2–6 vs 7–10) LH-RH agonist (⫾bicalutamide) Initial PSA (20 or less vs greater than 20 ng/ml) Testosterone breakthrough increases greater than 20 ng/dl Testosterone breakthrough increases greater than 50 ng/dl
Log Rank
p Value
3.27 0.01 6.71 0.99 0.10 1.78
0.0723 0.9095 0.0096 0.3188 0.7483 0.1819
7.74
0.0054
1,0
Cum survival free of PSA progression
1292
p 0.0207
0,8
Group 1
0,6
Group 2
0,4
0,2
Group 3
0,0 0
24
48
72
96
120
144
168
192 216
240
Months under ADT FIG. 1. Survival free of AIP according to serum testosterone behavior. Group 1, patients with all 3 serum testosterone determinations less than 20 ng/dl. Group 2, patients with breakthrough increases between 20 and 50 ng/dl. Group 3, patients with breakthrough increases greater than 50 ng/dl.
determinations less than 20 ng/dl), 90 months (95% CI 73–108) in group 2 (any increase between 20 and 50 ng/dl) and 72 months (95% CI 33–111) in group 3 (any increase greater than 50 ng/dl). Because patients with all 3 determinations less than 20 ng/dl (group 1) had a longer survival free of AIP than the other 2 groups, the clinically significant threshold of testosterone was investigated. The lowest serum testosterone threshold that was able to significantly distinguish groups related with the survival free of AIP was 32 ng/dl, p ⫽ 0.0258 (fig. 2). Patients with all 3 determinations of serum testosterone less than 32 ng/dl had a mean survival free of AIP of 137 (95% CI 104 –170) vs 88 months (95% CI 55–121) for those with any breakthrough increase greater than 32 ng/dl, p ⬍0.03. Relationship Between Breakthrough Increases and ADT Modality A cross tabulation analysis between ADT modality and breakthrough increases of serum testosterone greater than any threshold (20, 32 or 50 ng/dl) revealed no significant relation between treatment with bicalutamide and the incidence of breakthrough increases in serum testosterone. In the subset of patients on maximal androgen blockade the rate of breakthrough increases of serum testosterone greater than 50 ng/dl was 75.6% (33 of 45), similar to the 75.0% (21 of 28) in those patients treated with medical castration, p ⫽ 0.957. The clinical impact of ADT modality in those in whom breakthrough increases of serum testosterone developed at some point was analyzed. Taking the conventional threshold of 50 ng/dl, mean survival free of PSA progression was 115 months (95% CI 53–172) when receiving maximal androgen blockade as opposed to 32 months (95% CI 19 – 45) when receiving LH-RH agonist alone (p ⫽ 0.0249, fig. 3, A).
BREAKTHROUGH INCREASES OF TESTOSTERONE IN PROSTATE CANCER
Cumulate survival free of AIP (%)
1,0
p = 0.0258
0,8
1293
with medical castration has been established. In recent years the availability of new and more accurate assay methods has enabled the detection of serum testosterone levels below the threshold of 20 ng/dl,9 which is much lower than the conventional 50 ng/dl castration level defined after the phase III study for the Food and Drug Administration licensing of 22.5 mg leuprolide depot.13 Using this more strin-
Group 1
0,6
0,4
Group 2
0,2
0,0 0
50
100
150
200
250
Follow up (months) FIG. 2. Survival free of AIP analysis according to serum testosterone behavior in relation to lowest castrate threshold established at 32 ng/dl. Group 1, patients with all 3 serum testosterone determinations less than 32 ng/dl. Group 2, patients with breakthrough increases greater than 32 ng/dl.
Taking the 32 ng/dl threshold, mean survival free of AIP for patients on mAb was 98 months (95% CI 57–138) while it was 63 months (95% CI 39 – 87) for patients receiving LH-RH agonist alone, which did not reach significance (p ⫽ 0.3350). In patients without breakthrough increases greater than 32 ng/dl mean survival was 126 (95% CI 86 –167) (mAb) and 104 months (95% CI 77–132) (LH-RH agonist alone) (p ⫽ 0.2801), respectively. DISCUSSION This study demonstrates that breakthrough increases of serum testosterone in medically castrated patients with prostate cancer are not only frequent but also have clinical implications regarding PSA progression. In addition, the results suggest that 32 ng/dl might well be the lowest serum testosterone threshold with a clinical impact in these patients. Patients with breakthrough increases in serum testosterone greater than 32 ng/dl had a significantly shorter survival free of AIP (88 months) than those patients without these events (137 months). Finally, observations in the report imply that, at least for the highest and classic range of testosterone of 50 ng/dl, maximal androgen blockade (by adding bicalutamide to LH-RH agonist) might have a beneficial role. To our knowledge this is the first report to establish a direct relationship between testosterone increases and androgen independent progression. The importance of effective testosterone suppression is frequently emphasized on the basis that ineffective androgen suppression may result in increased prostate cancer mortality. However, this observation is based on more than 30-year-old studies with diethylstilbestrol.12 Currently no evidence of the clinical implications of breakthrough increases of testosterone when treated
FIG. 3. Survival free of AIP according to ADT modality and serum testosterone threshold established at 50 ng/dl. Group 1, patients treated with LH-RH agonist plus bicalutamide. Group 2, patients treated with LH-RH agonist alone. A, patients with breakthrough increases of serum testosterone greater than 50 ng/dl. B, patients without breakthrough increases.
1294
BREAKTHROUGH INCREASES OF TESTOSTERONE IN PROSTATE CANCER
gent definition of 20 ng/dl, recent publications have shown that several patients treated with current LH-RH agonist formulations do not achieve this castration level, and show breakthroughs and acute-on-chronic responses.11 In addition, up to 13% of men on depot LH-RH agonists will have testosterone values between 20 and 50 ng/dl.14 This has triggered the question of whether achieving testosterone levels less than 50 ng/dl corresponds to optimal castration, and if 20 ng/dl, based on the level obtained with orchiectomy as a benchmark, would be a better threshold for defining an appropriate castration and treatment response.6 The present study analyzes 3 consecutive serum testosterone determinations performed at 6-month intervals in medically castrated patients with prostate cancer. The rate of patients with all 3 serum testosterone determinations greater than 20 ng/dl was 43.8%, whereas breakthrough increases of serum testosterone between 20 and 50 ng/dl were detected in 31.5% of patients, and greater than 50 ng/dl in 24.7%. In a recent review Tombal reported a rate of breakthrough increase in serum testosterone greater than 20 ng/dl in 13% to 35% of patients treated with LH-RH agonists, and greater than 50 ng/dl in 2% to 13% of these patients.11 Of note, whether this discrepancy in hormonal levels translates into a therapeutic effect and a survival advantage or disadvantage is unknown. In fact, clinical trials comparing outcomes between the various forms of hormonal therapy have reported no survival advantage for orchiectomy that achieves lower testosterone levels than monthly LH-RH agonist (3.6 mg goserelin acetate).4,15 However, this must be interpreted cautiously because of a relatively short median followup (less than 15 months), and because neither the British nor the United States prostate study groups were significantly powered to detect small differences in outcome.4,15 In the present report we failed to demonstrate a direct clinical benefit of the lower cutoff value of 20 ng/dl. However, 32 ng/dl, which is considerably lower than 50 ng/dl, showed a significant impact on survival free of AIP. Of note, in classic report by Oefelein et al in which the idea of the lower the better was postulated, values of up to 30 ng/dl were also observed in surgically castrated cases.6 Central to the debate regarding therapeutic castration is the potential beneficial role of maximal androgen blockade compared to castration alone. A modest overall and cancer specific survival has been shown on meta-analysis but associated with increased adverse events and reduced quality of life.16 The National Comprehensive Cancer Network recommends that if the LH-RH agonist therapy testosterone reaches castration values less than 20 ng/dl, not to add an antiandrogen. However, if the serum testosterone level is greater than 20 ng/dl and the patient refuses surgical castration, the addition of an antiandrogen should be considered.17 These guidelines are recommendations but they have not been clinically confirmed.15 Observations in our report support a potential protective effect of the addition of bicalutamide. Time to AIP was longer in patients on mAb compared to those treated with LH-RH analogues alone, which was the case for those with (98 vs 63 months) or without (126 vs 104 months) breakthrough increases in testosterone greater than 32 ng/dl. However, this finding only reached statistical significance when using the classic 50 ng/dl threshold (115 vs 32 months, p ⫽ 0.0249).
Collectively the findings of the current report prove that testosterone breakthrough increase has prognostic value in medically castrated patients based on the direct evidence of longer AIP. Our findings establish for the first time a cutoff value, 32 ng/dl, with significant clinical impact because increases greater than this threshold were related to a lower survival free of AIP. Our data suggest a rationale for the use of maximal androgen blockade with bicalutamide when breakthrough increases of serum testosterone greater than 50 ng/dl are detected, which is in accordance with the National Comprehensive Cancer Network guidelines and based on possible clinical benefit in these patients. CONCLUSIONS Because physicians today are not routinely evaluating the testosterone level when initiating and following ADT, we suggest that routine measurement of serum testosterone should become part of clinical practice when evaluating the effects of hormonal therapy. A reasonable option to detect these breakthrough increases would be to monitor testosterone levels at PSA determination. Apart from having prognostic implications, knowing the level of testosterone would allow redosing the LH-RH agonist accordingly and considering maximal androgen blockade. A prospective, randomized and carefully designed trial contemplating clinical progression and specific mortality as the primary end point would be required to confirm these findings and reassess the cutoff level.
Abbreviations and Acronyms ADT AIP LH-RH mAb PCa PSA
⫽ ⫽ ⫽ ⫽ ⫽ ⫽
androgen deprivation therapy androgen independent progression luteinizing hormone releasing hormone maximal androgen blockade prostate cancer prostate specific antigen
REFERENCES 1.
2.
3.
4.
5. 6.
7.
Huggins C and Hodges C: Studies on prostate cancer: the effect of castration, estrogen and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate. Cancer Res 1941; 1: 293. Mongiat-Artus P and Teillac P: Role of luteinising hormone releasing hormone (LHRH) agonist in the management of prostate cancer. Eur Urol, suppl, 2005; 4: 4. Schally AV, Kastin AJ and Arimura A: Hypothalamic folliclestimulating hormone (FSH) and luteinizing hormone (LH)regulating hormone: structure, physiology and clinical studies. Fertil Steril 1971; 22: 703. Kaisary AV, Tyrrell CJ, Peeling WB and Griffiths K: Comparison of LHRH analogue (Zoladex) with orchiectomy in patients with metastatic prostatic carcinoma. Br J Urol 1991; 67: 502. The Leuprolide Study Group: Leuprolide versus diethylstilbestrol for metastatic prostate cancer. N Engl J Med 1984; 311: 1281. Oefelein MG, Feng A, Scolieri MJ, Ricchiutti D and Resnick MI: Reassessment of the definition of castrate levels of testosterone: implications for clinical decision making. Urology 2000; 56: 1021. Sharifi R and Browneller R: Serum testosterone suppression and potential for agonistic stimulation during chronic treat-
BREAKTHROUGH INCREASES OF TESTOSTERONE IN PROSTATE CANCER ment with monthly and 3-month depot formulations of leuprolide acetate for advanced prostate cancer. J Urol 2002; 168: 1001. 8. Wilke TJ and Utley DJ: Total testosterone, free-androgen index, calculated free testosterone, and free testosterone by analog RIA compared in hirsute women and otherwisenormal women with altered binding of sex-hormonebinding globulin. Clin Chem 1987; 33: 1372. 9. Wheeler MJ, D’Souza A, Matadeen J and Croos P: Ciba Corning ACS:180 testosterone assay evaluated. Clin Chem 1996; 42: 1445. 10. Zherdev AV, Byzova NA, Izumrudov VA and Dzantiev BB: Rapid polyelectrolyte-based immunofiltration technique for testosterone detection in serum samples. Analyst 2003; 128: 1275. 11. Tombal B: Appropriate castration with luteinising hormone releasing hormone (LHRH) agonist: what is the optimal level of testosterone. Eur Urol, suppl, 2005; 4: 14. 12. Jordan WP Jr, Blackard CE and Byar DP: Reconsideration of orchiectomy in the treatment of advanced prostatic carcinoma. South Med J 1977; 70: 1411.
13.
1295
Sharifi R, Bruskewitz RC, Gittleman MC, Graham SD Jr, Hudson PB and Stein B: Leuprorelin acetate 22.5 mg 12week depot formulation in the treatment of patients with advanced prostate cancer. Clin Ther 1996; 18: 647. 14. Oefelein MG and Cornum R: Failure to achieve castrate levels of testosterone during luteinizing hormone releasing hormone agonist therapy: the case for monitoring serum testosterone and a treatment decision algorithm. J Urol 2000; 164: 726. 15. Vogelzang NJ, Chodak GW, Soloway MS, Block NL, Schellhammer PF, Smith JA Jr et al: Goserelin versus orchiectomy in the treatment of advanced prostate cancer: final results of a randomized trial. Zoladex Prostate Study Group. Urology 1995; 46: 220. 16. Schmitt B, Bennett C, Seidenfeld J, Samson D and Wilt T: Maximal androgen blockade for advanced prostate cancer. Cochrane Database Syst Rev 2000; 2: CD001526. 17. Millikan R and Logothetis C: Update of the NCCN guidelines for treatment of prostate cancer. Oncology (Williston Park) 1997; 11: 180.