Fax (011) 39 02 2643 7298
Preoperative hypogonadism is not an independent predictor of high-risk disease in patients undergoing radical prostatectomy†
Article first published online: 1 MAR 2011
Copyright © 2011 American Cancer Society
Volume 117, Issue 17, pages 3953–3962, 1 September 2011
How to Cite
Salonia, A., Gallina, A., Briganti, A., Abdollah, F., Suardi, N., Capitanio, U., Colombo, R., Freschi, M., Rigatti, P. and Montorsi, F. (2011), Preoperative hypogonadism is not an independent predictor of high-risk disease in patients undergoing radical prostatectomy. Cancer, 117: 3953–3962. doi: 10.1002/cncr.25985
The authors thank Dragonfly Editorial for reviewing the language in the manuscript.
- Issue published online: 19 AUG 2011
- Article first published online: 1 MAR 2011
- Manuscript Accepted: 4 NOV 2010
- Manuscript Revised: 3 NOV 2010
- Manuscript Received: 29 JUN 2010
- high risk;
- prostate cancer;
- radical prostatectomy;
- pathologic stage;
- pathologic grade
A study was undertaken to assess the association between either preoperative serum total testosterone (TT) or hypogonadism (defined as TT <3 ng/mL) with high-risk prostate cancer (PCa) (defined as patients with pathological extracapsular extension [ECE], seminal vesicle invasion [SVI], or Gleason grades ≥4 + 3 [high-grade PCa]) at radical prostatectomy (RP).
A cohort of 673 consecutive Caucasian-European patients who underwent RP at a single institute was used. None of the patients had taken any hormonal neoadjuvant treatment or other hormonal preparations during the previous 12 months. Serum TT was measured the day before surgery (8-10 AM) in all cases. Logistic regression models tested the associations among predictors (eg, prostate-specific antigen, clinical stage, biopsy Gleason sum, body mass index, and TT) and ECE, SVI, and high-grade PCa.
Median TT was 4.5 ng/mL (mean, 4.5; range, 0.02-13.6). Hypogonadism was found in 144 (21.4%) patients, and severe hypogonadism (defined as TT <1 ng/mL) was observed in 37 (5.5%) men. Extracapsular extension, SVI, and high-grade PCa were found in 96 (14.6%), 88 (13.1%), and 153 (22.7%) patients, respectively. Patients with high-risk PCa had median TT comparable to those with low-risk disease. At multivariate analysis, TT did not achieve independent predictor status for ECE, SVI, and high-grade PCa. Only circulating TT <1 ng/mL was an independent predictor of SVI (odds ratio, 3.11; P = .006).
In contrast with previous reports, preoperative circulating TT levels were not associated with high-risk PCa. Likewise, hypogonadism did not achieve independent predictor status for high-risk PCa. Cancer 2011;. © 2011 American Cancer Society.
Sex hormones have been implicated in prostate carcinogenesis, through their roles in modulating cell proliferation, differentiation, and growth by altering the balance between cell proliferation and apoptosis.1-5 Androgens are vital for growth and maintenance of the prostate6; however, an association between circulating androgens and prostate cancer (PCa) has not been clearly confirmed in epidemiologic studies.6 Likewise, data from several population-based studies failed to show a significant association between circulating levels of androgens (including total testosterone [TT]) and increased risk of PCa.5-11
Conversely, low serum TT levels, rather than high levels, at time of PCa diagnosis have been found to be associated with advanced or high-grade PCa.3, 6, 12, 13 Animal studies showed that prolonged exposure to low androgen levels (ie, 10-fold lower than normal) resulted in a marked acceleration of prostate tumorigenesis, relative to cases exposed to androgen levels within the reference range.14 The exposure to reduced androgen levels may promote prostate tumorigenesis by selecting for molecular events that induce more aggressive, hormone-refractory tumors.14
Data from recent clinically localized PCa series treated with radical prostatectomy (RP) suggested that preoperative circulating TT levels were associated with advanced pathologic stages15-18; controversies still exist regarding the relationship between serum TT levels and Gleason score, a well-established proxy of cancer aggressiveness.6, 19, 20
These observations prompted us 1) to analyze the association between preoperative serum TT and high-risk disease (ie, extracapsular extension [ECE], seminal vesicle invasion [SVI], and high-grade PCa [defined as Gleason grades ≥4 + 3]) in a large cohort of nonscreened Caucasian-European PCa patients undergoing retropubic RP (RRP) at a single institution and 2) to evaluate the association between hypogonadism and high-risk disease.
MATERIALS AND METHODS
The analyses were based on a cohort of 780 consecutive nonscreened Caucasian-European patients who underwent RRP at a single academic referral center between June 2006 and June 2008. None of the patients had uncontrolled diabetes, thyroid disease, hyperprolactinemia, hypoalbuminemia, or liver diseases. For the specific purpose of this study, none of the patients had taken any hormonal neoadjuvant treatment or other hormonal preparations during the previous 12 months. Symptoms of late onset hypogonadism were not specifically collected for this cohort of patients.
All patients were comprehensively assessed with a detailed preoperative evaluation, including total serum prostate-specific antigen (PSA) (Abbott Axym PSA assay; Abbott Laboratories, Abbott Park, Ill); measured body mass index (BMI), defined as weight in kilograms divided by height in square meters; and clinical stage determined by a senior attending urologist (according to the 2002 American Joint Committee on Cancer staging system21).
We excluded 107 men because they missed 1 or more of the entry criteria (not comprehensive preoperative medical history [n = 29; 3.7%], missing or imprecise clinical stage [n = 39; 5.0%], missing measured preoperative BMI [n = 15; 1.9%], missing preoperative total PSA [n = 4; 0.5%], missing primary or secondary biopsy Gleason grades [n = 11; 1.4%], missing TT [n = 9; 1.2%]). A convenience sample of 673 (86.3%) patients was included in the analysis.
All RP specimens were fixed intact in 20% neural buffered formalin for 24 hours. The entire prostate and seminal vesicles were weighed and measured and then inked for surgical margin evaluation. The apical and bladder margins were then removed, bisected into right and left halves, sectioned sagittally, and submitted separately. The remainder of the gland was serially sectioned transversely across the urethra into approximately 3 to 4 mm slices from apex to base, sectioned into quadrants, and then submitted. The pathologic specimen assessment was performed by 1 dedicated genitourinary pathologist (M.F.).
For the specific purpose of this study and to reflect common practice of a clinical biochemistry laboratory, we elected to measure circulating TT using commercially available analytic methods. A single preoperative venous blood sample was drawn from each participant at least 4 weeks after transurethral ultrasound-guided prostate needle biopsies; samples were drawn after an overnight fast between 8 AM and 10 AM on the day before surgery, in compliance with the Clinical and Laboratory Standards Institute guidelines.22 In all cases, TT levels were measured via a direct chemiluminescence immunoassay (ADVIA Centaur; Siemens Medical Solutions Diagnostics, Deerfield, Ill). The same laboratory was used for all patients. The intra-assay and interassay coefficients of variation for TT were <6% and <8%, respectively.
Main Outcome Measures
The primary endpoint of the present study was to assess the association of TT and ECE (defined as presence of cancer outside the prostatic capsule and the seminal vesicles and the lymph nodes free of tumor),23 SVI (diagnosed if the tumor invaded the muscular wall of 1 or both seminal vesicles without evidence of lymph node invasion), or high-grade PCa at RRP. The secondary endpoint was to investigate whether preoperative hypogonadism (defined as TT <3 ng/mL)24 or severe hypogonadism (defined as TT <1 ng/mL) was significantly associated with high-risk PCa.
Data are presented as mean (median; range). The statistical significance of differences in means and proportions was tested with the 2-tailed t test and the chi-square2 test, respectively.
Logistic regression models tested the associations among predictors (eg, age, PSA, BMI, clinical stage, biopsy Gleason sum, TT, and either preoperative TT <3 ng/mL or TT <1 ng/mL) and ECE, SVI, and high-grade PCa. All statistical tests were performed using S-Plus Professional v.1 (MathSoft, Seattle, Wash). All tests were 2-sided, with a significance level set at .05.
The study was approved by the local ethics committee; likewise, the assay of this protocol was approved by the local institutional review board. Informed consent was obtained from all patients before enrollment. Patients agreed to deliver their own anonymous information for future studies.
Table 1 lists the characteristics and the descriptive statistics of the entire cohort of patients. ECE was found in 96 (14.6%) patients, SVI was found in 88 (13.1%) patients, and high-grade PCa was found in 153 (22.7%) patients. Tables 2, 3, and 4 detail ECE, SVI, and high-grade PCa status, respectively, for the whole cohort. Patients with ECE had significantly higher PSA than those without ECE. Likewise, a higher proportion of ECE patients had more advanced clinical stage and a lower rate of well-differentiated biopsy Gleason sum than men without ECE. Conversely, no significant differences were found between patients with or without ECE regarding age, BMI, and preoperative TT. Patients with SVI had significantly higher PSA than those without SVI. Similarly, a higher proportion of SVI patients had more advanced clinical stage and a higher rate of biopsy Gleason sum ≥7 than those without SVI. Conversely, no significant differences were found between patients with or without SVI regarding age, BMI, and preoperative TT. Patients with high-grade PCa had significantly higher total PSA than those without high-grade PCa. A higher proportion of high-grade PCa patients had more advanced clinical stage and a higher rate of less differentiated tumors at biopsy than those without high-grade PCa. No significant differences were found between patients with or without high-grade PCa regarding age, BMI, and preoperative TT.
|No. of patients||673|
|Mean [median]||64.0 [64.6]|
|Mean [median]||26.2 [25.9]|
|Mean [median]||11.6 [6.7]|
|Mean [median]||4.5 [4.5]|
|Clinical stage, No. (%)|
|Biopsy Gleason sum, No. (%)|
|Pathological stage, No. (%)|
|Pathological Gleason sum, No. (%)|
|Patients, No. (%)||96 (14.3)||577 (85.7)|
|Age, y; mean [median]||63.9 [64.7]||64.1 [64.6]||.84||−1.39 to 1.71|
|BMI, kg/m2; mean [median]||26.6 [26.4]||26.2 [25.7]||.23||−1.08 to 0.26|
|PSA, ng/mL; mean [median]||20.5 [7.7]||10.1 [6.7]||.004||−17.34 to −3.42|
|TT, ng/mL; mean [median]||4.73 [4.67]||4.49 [4.41]||.3||−0.69 to 0.21|
|Clinical stage, No. (%)|
|T1c||47 (49.0)||359 (62.2)||.02||5.45||2.45 to 23.95|
|T2||30 (31.3)||161 (27.9)||.57||0.32||−6.57 to 13.37|
|T3||19 (19.7)||57 (9.9)||.008||6.94||1.48 to 18.12|
|Biopsy Gleason sum, No. (%)|
|≤6||45 (46.9)||387 (67.1)||.0002||13.75||9.51 to 30.89|
|7||39 (40.6)||146 (25.3)||.003||8.92||4.86 to 25.74|
|≥8||12 (12.5)||44 (7.6)||.16||1.99||−2.06 to 11.86|
|Patients, No. (%)||88 (13.1)||585 (86.9)|
|Age, y; mean [median]||65.0 [65.4]||63.8 [64.6]||.16||−1.14 to 0.82|
|BMI, kg/m2; mean [median]||26.3 [25.7]||26.2 [25.9]||.72||−0.82 to 0.57|
|PSA, ng/mL; mean [median]||24.2 [10.3]||10.2 [6.6]||<.001||−21.70 to −7.28|
|TT, ng/mL; mean [median]||4.41 [4.68]||4.54 [4.43]||.61||−0.34 to 0.59|
|Clinical stage, No. (%)|
|T1c||33 (37.5)||373 (63.8)||<.0001||21.03||15.46 to 37.14|
|T2||27 (30.7)||164 (28.0)||.69||0.16||−7.60 to 13.00|
|T3||28 (31.8)||48 (8.2)||<.0001||40.23||13.62 to 33.58|
|Biopsy Gleason sum, No. (%)|
|≤6||19 (21.6)||413 (70.6)||<.0001||77.79||39.64 to 58.36|
|7||42 (47.7)||143 (24.4)||<.0001||19.70||12.30 to 34.30|
|≥8||27 (30.7)||29 (5.0)||<.0001||62.63||15.90 to 35.50|
|Patients, No. (%)||153 (22.7)||520 (77.3)|
|Age, y; mean [median]||64.5 [65.2]||63.9 [64.6]||.41||−1.84 to 0.74|
|BMI, kg/m2; mean [median]||26.6 [26.1]||26.1 [25.6]||.065||−1.09 to 0.03|
|PSA, ng/mL; mean [median]||24.2 [8.25]||7.9 [6.4]||<.001||−21.95 to −10.50|
|TT, ng/mL; mean [median]||4.39 [4.21]||4.56 [4.50]||.37||−0.20 to 0.55|
|Clinical stage, No. (%)|
|T1c||64 (41.8)||342 (65.8)||<.0001||27.46||15.18 to 32.82|
|T2||48 (31.4)||143 (27.5)||.40||0.70||−4.40 to 12.20|
|T3||41 (26.8)||35 (6.7)||<.0001||45.77||12.76 to 27.44|
|Biopsy Gleason sum, No. (%)|
|≤6||48 (31.4)||384 (73.8)||<.0001||90.59||34.13 to 50.67|
|7||59 (38.6)||126 (24.2)||.0007||11.59||5.85 to 22.95|
|≥8||46 (30.1)||10 (1.9)||<.0001||119.70||20.84 to 35.56|
Hypogonadism was found in 144 (21.4%) patients, and severe hypogonadism was observed in 37 (5.5%) men. Tables 5 and 6 stratify patients with either ECE, SVI, or high-grade PCa according eugonadism, hypogonadism, and severe hypogonadism status, respectively. A higher proportion of SVI and high-grade PCa patients were found both in the hypogonadal and severely hypogonadal group than in eugonadal men (all P ≤ .001). No significant differences were found regarding ECE according to the preoperative circulating TT values.
|Variable||TT ≥3 ng/mL [Eugonadism], No. (%)||TT <3 ng/mL [Hypogonadism], No. (%)||Pa||Chi-Square||95% CI|
|Patients||529 (78.6)||144 (21.4)|
|ECE||76 (14.4)||20 (13.9)||.99||0.00||−5.89 to 6.89|
|SVI||58 (11.0)||30 (20.8)||.003||8.71||2.65 to 16.95|
|HGPCa||105 (19.8)||48 (33.3)||.0009||11.00||5.09 to 21.91|
|Variable||TT ≥3 ng/mL [Eugonadism], No. (%)||TT <1 ng/mL [Severe hypogonadism], No. (%)||Pa||Chi-Square||95% CI|
|Patients||529 (78.6)||37 (6.5)|
|ECE||76 (14.4)||7 (18.9)||.61||0.26||−8.47 to 17.47|
|SVI||58 (11.0)||17 (45.9)||<.0001||33.60||18.62 to 51.18|
|HGPCa||105 (19.8)||22 (59.5)||<.0001||29.12||23.52 to 55.88|
According to univariate analysis (Table 7), clinical stage and biopsy Gleason sum were significantly associated with ECE. Likewise, multivariate analysis confirmed that clinical stage and biopsy Gleason sum achieved independent predictor status for ECE. Conversely, no significant associations were found between ECE and the other possible predictors.
|OR; P||TT, OR; P||TT <3 ng/mL, OR; P||TT <1 ng/mL, OR; P|
|Age||1.00; .83||0.99; .62||0.99; .62||1.00; .78|
|BMI||1.04; .23||1.07; .08||1.06; .08||1.06; .08|
|PSA||1.01; .05||1.00; .70||1.00; .72||1.01; .14|
|TT||1.06; .29||1.06; .26||—;—||—;—|
|TT <3 ng/mL||0.96; .88||—;—||1.23; .42||—;—|
|TT <1 ng/mL||1.43; .41||—;—||—;—||0.77; .59|
|Clinical stage||—; .009||—; .04||—; .038||—; .07|
|T2 vs T1c||1.42; .16||1.34; .26||1.34; .26||1.37; .23|
|T3 vs T1c||2.55; .002||2.30; .01||2.33; .01||2.17; .02|
|Biopsy Gleason sum||—; .001||—; .005||—; .004||—; .006|
|7 vs ≤6||2.30; .001||2.18; .001||2.27; .001||2.19; .001|
|≥8 vs ≤6||2.36; .02||1.73; .16||2.27; .001||1.55; .29|
Total PSA, hypogonadism, clinical stage, and biopsy Gleason sum were significantly associated with SVI (all P ≤ .02) (Table 8). Conversely, multivariate analysis confirmed that only TT <1 ng/mL, clinical stage, and biopsy Gleason sum achieved independent predictor status for SVI. No significant associations were found between SVI and age, BMI, PSA, or TT.
|OR; P||TT, OR; P||TT <3 ng/mL, OR; P||TT <1 ng/mL, OR; P|
|Age||1.02; .16||1.01; .49||1.01; .49||1.01; .46|
|BMI||1.06; .88||1.04; .30||1.03; .47||1.04; .35|
|PSA||1.01; .01||1.00; .46||1.00; .54||1.00; .76|
|TT||0.97; .61||0.96; .47||—;—||—;—|
|TT <3 ng/mL||2.14; .002||—;—||1.55; .14||—;—|
|TT <1 ng/mL||6.76; <.001||—;—||—;—||3.11; .006|
|Clinical stage||—; <.001||—; .001||—; .002||—; .003|
|T2 vs T1c||1.9; .02||1.36; .31||1.38; .29||1.40; .27|
|T3 vs T1c||6.6; <.001||3.51; <.001||3.44; <.001||3.36; .001|
|Biopsy Gleason sum||—; <.001||—; <.001||—; <.001||—; .003|
|7 vs ≤6||6.4; <.001||5.80; <.001||5.62; <.001||5.26; <.001|
|≥8 vs ≤6||20.2; <.001||13.14; <.001||12.39; <.001||11.34; <.001|
At univariate analysis, PSA, clinical stage, biopsy Gleason sum, and hypogonadism were significantly associated with high-grade PCa (all P ≤ .007) (Table 9). Conversely, multivariate analysis found that only PSA, BMI, clinical stage, and biopsy Gleason sum achieved independent predictor status for high-grade PCa. In contrast, no significant associations were found between high-grade PCa and age, TT, or hypogonadism.
|OR; P||TT, OR; P||TT <3 ng/mL, OR; P||TT <1 ng/mL, OR; P|
|Age||1.01; .41||1.00; .98||1.00; .99||1.00; .97|
|BMI||1.44; .25||1.11; .003||1.11; .004||1.11; .002|
|PSA||1.06; <.001||1.04; <.001||1.04; .001||1.04; .001|
|TT||0.96; .37||0.96; .48||—;—||—;—|
|TT <3 ng/mL||2.02; .001||—;—||1.24; .42||—;—|
|TT <1 ng/mL||5.65; <.001||—;—||—;—||1.74; .23|
|Clinical stage||—; <.001||—; .002||—; .002||—; .002|
|T2 vs T1c||1.8; .007||1.33; .25||1.34; .24||1.34; .24|
|T3 vs T1c||6.3; <.001||3.22; <.001||3.21; <.001||3.20; <.001|
|Biopsy Gleason sum||—; <.001||—; <.001||—; <.001||—; <.001|
|7 vs ≤6||3.7; <.001||3.28; <.001||3.27; <.001||3.16; <.001|
|≥8 vs ≤6||36.8 <.001||24.07; <.001||23.38; <.001||23.12; <.001|
We tested whether preoperative circulating TT was associated with high-risk disease in a large cohort of nonscreened Caucasian-European PCa patients undergoing RRP at a single academic institute. Our interest was fueled by the existing controversies regarding the role of circulating TT as an established predictor of pathologic outcomes at RP and the lack of data dedicated to a nonscreened, homogeneous cohort of Caucasian-European PCa patients. Our findings show that immediately preoperative serum TT is not a multivariate predictor of either locally advanced or less differentiated prostate tumors.
According to the widely used biochemical definition of hypogonadism (ie, serum TT levels <3 ng/mL12, 24), the prevalence of hypogonadism was 21.4% within our cohort of patients and is comparable to prevalence estimates from other studies.25, 26 Higher rates of both SVI and high-grade PCa were found among preoperatively hypogonadal and severely hypogonadal men as compared with men with physiologic circulating TT levels. Moreover, current findings showed that both serum TT levels <3 ng/mL and <1 ng/mL were univariate predictors of both SVI and high-grade PCa; in contrast, only TT <1 ng/mL values reached multivariate independent predictor status for SVI.
One strength is that the current study was a single-institute survey with a large cohort of nonscreened, homogeneous, same-race patients, for which all surgical procedures and specimen evaluations were performed using a consistent method. A second strength was that the patients included in this study presented a wide variety of low- and high-risk tumors, possibly allowing for adequate variability in the baseline serum TT values to provide robust analysis. Third, all blood samples were correctly drawn after an overnight fast between 8 AM and 10 AM (level 2a, grade A27), thus avoiding a potential methodological flaw because of different collection times and diurnal variation of the steroid hormones. A further strength of the current analysis is that we rigorously excluded men with uncontrolled diabetes, thyroid disease, hyperprolactinemia, hypoalbuminemia, or liver diseases as well as those patients who had taken any form of hormonal preparation during the previous 12 months.28
Although contradictory findings have been published, androgens are still believed to be critical determinants in normal and neoplastic growth and development of the prostate.11 However, substantial controversy exists regarding the association between pretreatment T and pathologic disease outcome in patients with PCa.10, 14-16, 20 Massengill et al,15 in a large retrospective revision of clinical records of patients who underwent RP, showed that those with nonorgan-confined PCa had significantly lower pretreatment TT levels compared with those with organ-confined cancer. Likewise, pretreatment TT emerged as a significant multivariate independent predictor of extraprostatic disease within the same cohort of men.15 Similarly, in a small cohort of subjects with clinically organ-confined tumors who underwent RP, Imamoto et al16 reported that pretreatment T level was significantly lower in patients with nonorgan-confined PCa. At multivariate analysis, pretreatment serum T level was found to be a significant predictor of extraprostatic disease.16 In the retrospective analysis conducted by Isom-Batz et al,17 lower T values also significantly correlated with adverse pathologic stage on multivariate analysis, as did clinical stage, biopsy grade, and PSA. Our findings only showed an univariate association between preoperative circulating TT values and the pathologic features at RP; in contrast, TT did not acquire independent predictor status for the same outcomes.
A few studies have assessed the relationship between systemic sex steroids, regardless of the biochemical condition of hypogonadism, and PCa aggressiveness and show conflicting data.11, 29 Ide et al,20 using prostate biopsies obtained from men with abnormal PSA levels, found that circulating T values were significant predictors of high-grade PCa. Similarly, Pierorazio et al30 recently observed that higher free T (FT) index was associated with an increased risk of aggressive PCa (hazard ratio, 2.07) among men aged >65 years. In contrast, several surveys did not show any association between preoperative TT levels and Gleason grades, either in series limited to patients with clinically localized PCa at diagnosis6, 15-17 or in a more clinically heterogeneous cohort of PCa patients.10, 31 Our results support these latter findings, strengthened by a large, homogenous cohort of Caucasian-European men with a broad spectrum of clinical presentations and surgical pathologic specimens.
A larger amount of data has been published regarding the association between circulating low TT levels (ie, biochemical hypogonadism) and cancer aggressiveness. Hoffman et al32 showed that patients with newly diagnosed PCa and low FT had more extensive disease at biopsy. In addition, all men with a biopsy Gleason score ≥8 had low serum FT.32 Similarly, Schatzl et al33 found that patients with high Gleason scores have lower serum levels of T, 17β-estradiol, and gonadotropins. Massengill et al15 reported that as T decreases, patients have an increased likelihood of nonorgan-confined disease. Interestingly, a prospective case-cohort study of 17,049 men, including 524 cases diagnosed during a mean follow-up of 8.7 years and a randomly sampled subcohort of 1859 men, showed that high levels of circulating T and adrenal androgens were associated with reduced risk of aggressive PCa but not with reduced risk of nonaggressive disease.19 Our findings showed that hypogonadism did not achieve independent predictor status for high-risk tumors, considering ECE, SVI, and high-grade PCa. Only severely low serum TT was an independent predictor of SVI. To the best of our knowledge, no previously published data confirm this finding in a homogenous cohort of nonscreened Caucasian-European PCa patients who underwent RRP at a single referral institute.
Our analysis found that preoperative BMI achieved independent predictor status for high-grade PCa. This result contrasted with previous observations of Gallina et al,34 who reported that among European men, obesity does not predispose to more aggressive PCa at biopsy and does not change the ability to identify patients who may harbor high-grade PCa at RP. Conversely, our findings fit with several studies that have shown that an elevated BMI significantly correlates with higher PCa grades.35, 36
Our study is not devoid of limitations. Although substantial controversy still exists regarding the association between pretreatment T and pathologic disease outcome in patients with PCa,10, 14-16, 20 our findings do not support previously published results showing that lower T values significantly correlated with adverse pathologic stage. In this context, it is important to highlight that although patients considered for these analyses were rigorously homogeneous for inclusion criteria, we cannot completely exclude the presence of unstudied biases.
A second limitation is that we lacked calculated FT, which may be derived from Vermeulen's formula,37 that generally reflects the clinical situation more accurately than total plasma hormone levels. Similarly, our study did not use gas chromatography–mass spectrometry,38 which is considered the gold standard for measuring circulating TT levels; in contrast, for the specific purpose of the study and to reflect common practice of a clinical biochemistry laboratory, we elected to measure circulating TT using commercially available analytic methods.
Another limitation is the lack of a complete functional assessment of the hypothalamic–pituitary–gonadal (HPG) hormone axis. Heracek et al39 reported that significantly lower serum levels of follicle-stimulating hormone were found in patients with localized PCa than in those with locally advanced tumors. Schatzl et al33 showed that gonadotropins were lower in parallel with steroid hormones in their cohort of PCa patients, thus suggesting a sort of tumor-mediated suppression of the HPG hormone axis, particularly in men with high-grade PCa. Furthermore, a single assessment of circulating androgen concentrations cannot correctly determine total androgenicity, because it reflects the function of the entire androgen signaling pathway, from the hypothalamus to the testis and back.40, 41 However, a single fasting morning venous blood sample makes the assessment of TT more user friendly in everyday clinical practice.
In addition, one may speculate that a single serum assessment might not adequately represent the prostate's hormonal environment throughout the life span of each subject or at least for the duration of the malignant transformation and progression toward higher aggressiveness. Notably, Miller et al42 found that serum T and gonadotropin levels were significantly increased after RP, whereas Lukkarinen et al43 did not get similar results in patients who underwent simple prostatectomy but not RP. Therefore, because a significant postoperative rise of circulating T levels has been demonstrated, especially in patients with low preoperative T levels, it might be hypothesized that PCa cells produce some substance that disturbs the pituitary–gonadal axis by either reducing or suppressing T production.44, 45 It is almost clinically impossible to follow the circulating hormonal milieu of a sufficient number of men throughout the entire life span to assess the role of sex steroids as independent predictors of the eventual biology of PCa; however, it seems reasonable to follow men who underwent curative treatment and to assess their hormones over time.6 Our current findings lack follow-up data but clearly show that immediately preoperative circulating TT does not achieve independent predictor status in pathologic terms.
Recent evidence supports the concept of limited contributions for intratumoral steroidogenesis in PCa.46 However, further complicating matters is the poor correlation between androgen concentrations in the circulation and actual concentrations in the prostate.47 We would consider it more significant to target complete T levels and thus consider both systemic and intracrine contributions to the prostatic androgen microenvironment.38, 48, 49 Although the lack of such elegant assessment might be considered a methodological flaw of our study, it certainly exceeds the clinical applicability of biochemical parameters that may be of interest as routinely available predictors of pathologic outcomes at RP.
Although we consider including a large general-population–based cohort of nonscreened, homogeneous, same-race patients to be a major strength of the current study, at a referral institute, we collected patients for whom biopsy technique as well as number of cores at biopsy and specimen evaluation were not always performed in a consistent way. A further limitation of our study is that we did not add either the percentage of cancer in biopsy specimens or the prostate tumor volume to the multivariate analysis, because both of these variables were previously reported to be highly correlated with pathologic outcomes at RP.50
In contrast to previously published studies, our data suggest that immediately preoperative circulating TT levels are not correlated with high-risk PCa in patients undergoing RRP. Likewise, preoperative hypogonadism per se was not an independent predictor of high risk PCa. Whether this difference may be considered simply a result of the different PCa patient populations in these studies rather than the absence of a biologically consistent correlation deserves major attention in prospective surveys. Further studies are needed to more comprehensively analyze the whole HPG hormone axis.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
- 21American Joint Committee on Cancer Staging Manual. 6th ed. New York, NY: Springer; 2002., , .
- 22How to Define, Determine, and Utilize Reference Intervals in the Clinical Laboratory; Approved Guidelines. Villanova, PA: National Committee for Clinical Laboratory Standards; 1995.
- 24American Association of Clinical Endocrinologists. Medical guidelines for clinical practice for the evaluation and treatment of hypogonadism in adult male patients—2002 update. Endocr Pract. 2002; 8: 440-456.