SEARCH

SEARCH BY CITATION

Keywords:

  • prostatic neoplasm;
  • prostatectomy;
  • testosterone;
  • hypogonadism;
  • pathological grade;
  • pathological stage

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Study Type – Prognosis (prospective cohort)

Level of Evidence 2b

What's known on the subject? and What does the study add?

Previous data from clinically localized prostate cancer (PCa) series treated with radical prostatectomy (RP) have suggested that low preoperative serum total testosterone level is associated with more aggressive PCa; however, the definition of low preoperative total testosterone level varied among these studies (from 220 ng/dL to 387 ng/dL). Moreover, no relevant data exist in the literature regarding ethnic Chinese patients.

The study shows that the most widely used threshold for low pretreatment total testosterone level (total testosterone < 300 ng/dL) is not appropriate for ethnic Chinese patients, because it could not distinguish patients with more aggressive PCa from those with less aggressive disease. Setting the threshold at the level of total testosterone < 250 ng/dL works better, because pretreatment total testosterone < 250 ng/dL is associated with a significantly higher incidence of Gleason score 8–10 disease in RP specimens.

OBJECTIVE

  • • 
    To investigate the relationship between preoperative serum total testosterone level and prognostic factors of Chinese patients with clinically localized prostate cancer (PCa).

PATIENTS AND METHODS

  • • 
    A total of 110 patients with localized PCa, treated by radical prostatectomy (RP), were included in this prospective study.
  • • 
    Clinical and pathological data from each patient were collected. Total testosterone was measured on the morning of surgery.
  • • 
    Total testosterone levels for each patient were compared using two thresholds: threshold 1 (total testosterone <300 ng/dL vs total testosterone ≥300 ng/dL) and threshold 2 (total testosterone <250 ng/dL vs total testosterone ≥250 ng/dL).

RESULTS

  • • 
    The median preoperative total testosterone level was 346 ng/dL. Gleason scores of ≤6, 7 and ≥8 were found in the RP specimens from 21 (19.1%), 67 (60.9%) and 22 (20.0%) patients, respectively.
  • • 
    Compared with those with low grade disease, patients with high grade disease (Gleason score ≥ 8) in RP specimens had a significantly lower preoperative total testosterone.
  • • 
    When comparing 35 patients with hypogonadism with 75 patients with eugonadism, classified by threshold 1, no significant relationships were found.
  • • 
    When comparing 18 patients with hypogonadism with 92 patients with eugonadism, classified by threshold 2, pathological Gleason score ≥ 8 tumours were more common in patients with hypogonadism.

CONCLUSION

  • • 
    Setting the threshold for hypogonadism at the level of pretreatment serum total testosterone <250 ng/dL is appropriate for ethnic Chinese patients with localized PCa, because patients with pretreatment total testosterone <250 ng/dL are associated with a higher incidence of Gleason score 8–10 disease in RP specimens.

Abbreviations
NA

not available

PCa

prostate cancer

RP

radical prostatectomy

BMI

body mass index

PCPT

Prostate Cancer Prevention Trial

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Androgens are vital for normal development and growth of the prostate [1], but the influence of androgens on prostate carcinogenesis is not well understood [1]. Previously, it was believed that high testosterone levels increase the risk of prostate cancer (PCa) and that low testosterone is protective [2]. Yet recently, evidence-based medicine has shown this is not true [2]. In the late 1990s, Morgentaler et al. [3] reported that a high prevalence of biopsy-detectable PCa was identified in men with low total or free testosterone levels. This was the first study to indicate that low testosterone was not protective against PCa [3]. In addition, extensive reviews have found no compelling evidence, in the PSA era, that men who had undergone testosterone replacement therapy were at increased risk of PCa [2,4], and data from several population-based studies have failed to show a significant association between high circulating levels of androgen and increased risk of PCa [5,6].

More interestingly, data from clinically localized PCa series treated with radical prostatectomy (RP) suggested that low preoperative serum total testosterone level was associated with more aggressive PCa, with a higher pathological stage, a higher rate of positive surgical margins or a higher Gleason score [7–12], but the definition of low preoperative total testosterone level varied among these studies (from 220 ng/dL to 387 ng/dL) [7–12]. In addition, most of these studies were performed in Caucasian patients and endocrinological studies have shown that there is significant ethnic variability in circulating total and free testosterone concentration [13,14]; currently available data, therefore, have limited value for ethnic Chinese patients.

To investigate the relationship between preoperative serum total testosterone level and prognostic factors of clinically localized PCa in ethnic Chinese patients, we conducted a prospective study in 110 patients with PCa who underwent open retropubic RP at our institution.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

PATIENTS

A total of 110 ethnic Chinese patients with clinically localized PCa, who underwent open retropubic RP at our institution between March 2011 and January 2012, were included in this prospective study. Data including patient age, weight, height, body mass index (BMI), clinical presentation, comorbidities, concomitant treatments, preoperative PSA level were collected. No patient received any kind of neoadjuvant hormonal therapy. Patients with uncontrolled diabetes, thyroid disease, hyperprolactinaemia, hypoalbuminaemia, or abnormal liver or renal function were excluded. All patients underwent RP at least 4 weeks after prostate biopsy. Some of the prostate biopsies were performed in outside hospitals, but all biopsy specimens were centrally reviewed by two dedicated genitourinary pathologists at our institution. Histological slides of all RP specimens were assessed by the same senior genitourinary pathologist throughout the study. The clinical and pathological stages were determined according to the 2010 American Joint Committee on Cancer TNM classification system. Histopathological grading of the biopsy and RP specimens was performed according to the Gleason score system. A positive surgical margin was defined as the presence of cancer at the inked surface of the resected specimen. The study was carried out in accordance with the ethical standards of the Helsinki Declaration II and approved by our institutional review board. Written informed consent was obtained from each patient before any study-specific investigation was carried out.

TESTOSTERONE ASSAY

Blood samples for total testosterone determination were collected after an overnight fast on the morning of prostatic surgery between 7:00 and 10:00 h, as recommended in the Endocrine Society guideline [15]. Total testosterone in all samples was detected in the same laboratory using Access® Testosterone (Beckman Coulter, Fullerton, CA, USA) based on an electrochemiluminescence immunoassay technique. The assay was used according to the manufacturer's recommendations.

Currently, there are no generally accepted lower limits of normal serum total testosterone [15]. In the present study, we used two common definitions for hypogonadism: (i) a threshold of total testosterone < 300 ng/dL; and (ii) a threshold of total testosterone < 250 ng/dL [16,17].

STATISTICAL ANALYSIS

Characteristics of the 88 patients with low grade disease (RP specimen Gleason score < 8) were compared with the remaining 22 patients with high grade disease (RP specimen Gleason score ≥ 8). Using the two different definitions for hypogonadism, 110 patients were classified into two groups: hypogonadism and eugonadism. Patients were compared by total testosterone levels: threshold 1 (total testosterone <300 ng/dL vs total testosterone ≥300 ng/dL) and threshold 2 (total testosterone <250 ng/dL vs total testosterone ≥250 ng/dL). Student's t-test was used to compare continuous variables. The Fisher exact (small size sample) or Pearson's chi-squared test were used to compare categorical variables. All statistical tests were two-sided and a P value <0.05 was considered to indicate statistical significance. SPSS 16.0 (SPSS, Chicago, IL, USA) was used for the statistical analysis.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Table 1 shows patient characteristics and baseline data. The median patient age was 67 years, the median preoperative PSA level was 11.9 ng/mL, and the median preoperative total testosterone level was 346 ng/dL. In RP specimens, Gleason score ≤ 6, 7 and ≥8 were observed in 21 (19.1%), 67 (60.9%) and 22 (20.0%) patients. The overall pathological stage was pT2 in 69 (62.7%) patients and pT3 in 41 (37.3%).

Table 1. Clinicopathological characteristics of 110 patients included in the study
CharacteristicsMean (median)Range
Age at surgery, year66.2 (67.0)49.0–78.0
Weight, kg67.6 (68.0)40.0–89.0
Height, m169.8 (170.0)155.0–183.0
BMI, kg/m223.5 (23.6)15.6–32.3
Preoperative PSA, ng/mL19.8 (11.9)1.4–169.1
Preoperative total testosterone, ng/dL393 (346)46–966
Medical history, n (%)  
 Hypertension38 (34.5) 
 Ischaemic heart disease12 (10.9) 
 Diabetes mellitus8 (7.3) 
Biopsy Gleason score, n (%)  
 ≤642 (38.2) 
 743 (39.1) 
 ≥825 (22.7) 
Clinical tumour stage, n (%)  
 T152 (47.3) 
 T241 (37.3) 
 T317 (15.4) 
Pathological Gleason score, n (%)  
 ≤621 (19.1) 
 767 (60.9) 
 ≥822 (20.0) 
Pathological tumour stage, n (%)  
 T2a–T2b9 (8.2) 
 T2c60 (54.6) 
 T3a19 (17.3) 
 T3b22 (20.0) 
Positive surgical margins, n (%)  
 Yes32 (29.1) 
 No78 (71.9) 
Positive pelvic lymph node, n (%)  
 Yes6 (5.5) 
 No104 (94.5) 

When comparing the 88 patients with low grade disease with the 22 with high grade disease in RP specimens, preoperative total testosterone level was lower in patients with high grade disease (P= 0.028, Table 2). Stage pT3a–T3b was more common in patients with high grade disease (P= 0.006, Table 2). No significant relationships were detected between the grade of disease and other variables.

Table 2. Comparison of characteristics of patients with low grade disease and those with high grade disease
VariablesLow grade disease (Pathological Gleason score < 8)High grade disease (Pathological Gleason score ≥ 8) P
No. of patients8822 
Mean (range) age, years66.4 (49.0–78.0)65.5 (54.0–77.0)0.583
Mean (range) weight, kg,67.2 (40.0–89.0)69.0 (54.0–84.0)0.356
Mean (range) height, m170.1 (155.0–183.0)168.4 (162.0–175.0)0.112
Mean (range) BMI, kg/m2,23.2 (15.6–32.3)24.3 (19.1–28.7)0.082
Mean (range) PSA, ng/mL19.4 (1.4–169.1)21.2 (2.0–96.9)0.790
Mean (range) total testosterone, ng/dL,410 (163–966)320 (46–761)0.028
Pathological stage, n (%)   
 T2a–T2c61 (69.3)8 (36.4)0.006
 T3a–T3b27 (30.7)14 (63.6) 
Positive surgical margins, n (%)   
 Yes28 (31.8)4 (18.2)0.295
 No60 (68.2)18 (81.8) 

Of the 110 patients, 35 (31.8%) were classified as having hypogonadism using the first threshold (total testosterone < 300 ng/dL) and 18 (16.4%) were classified as having hypogonadism using the second threshold (total testosterone < 250 ng/dL [Table 3]). When comparing the 35 patients with hypogonadism with the 75 patients with eugonadism, classified by the first threshold, no significant relationships were found between the total testosterone levels and all the variables. When comparing the 18 patients with hypogonadism with the 92 patients with eugonadism, classified by the second threshold, high grade cancer (pathological Gleason score ≥ 8) was found in 44.4% (8/18) patients with hypogonadism, which was significantly higher than in those patients with eugonadism (15.2%, 14/92). Moreover, BMI was also significantly higher in patients with hypogonadism (P= 0.041, Table 3).

Table 3. Comparison of characteristics of patients with hypogonadism vs eugonadism (classified using two different thresholds)
VariablesThreshold 1 (Hypogonadism: total testosterone <300 ng/dL)Threshold 2 (Hypogonadism: total testosterone <250 ng/dL)
HypogonadismEugonadism P HypogonadismEugonadism P
No. of patients3575 1892 
Mean (range) age, years65.8 (49–77)66.4 (56–78)0.61865.7 (49–77)66.3 (52–78)0.721
Mean (range) BMI, kg/m223.8 (15.6–29.4)23.3 (15.6–32.3)0.56224.2 (15.6–32.3)23.1 (15.6–28.7)0.041
Mean (range) PSA, ng/mL26.8 (2.0–169.1)17.5 (1.4–122.1)0.22117.1 (1.4–66.4)20.3 (2.0–169.1)0.662
Pathological Gleason score, n (%)      
 <825 (71.4)63 (84.0)0.12510 (55.6)78 (84.8)0.005
 ≥810 (28.6)12 (16.0) 8 (44.4)14 (15.2) 
Pathological stage, n (%)      
 T2a–T2c19 (54.3)50 (66.7)0.21110 (55.6)59 (64.1)0.491
 T3a–T3b16 (45.7)25 (33.3) 8 (44.4)33 (35.9) 
Positive surgical margins, n (%)      
 Yes6 (17.1)26 (34.7)0.0592 (11.1)30 (32.6)0.066
 No29 (82.9)49 (65.3) 16 (88.9)62 (67.4) 

Table 4 [7–12,17–19] has a summary of the present data and recently published studies that support correlations between low pretreatment total testosterone and poor prognostic factors in patients with localized PCa treated by RP. It should be noted that the definition of low pretreatment total testosterone varied among these studies, although 300 ng/dL was the most commonly used threshold.

Table 4. Summary of studies supporting correlations between low pretreatment total testosterone and poor prognostic factors in patients with localized PCa
AuthorCountry where study conductedSamplesThreshold for low pretreatment total testosterone (ng/dL)Patients with low pretreatment total testosterone (%)Pretreatment total testosterone level (ng/dL)Correlated prognostic factors
MeanMedianRange
Present studyChinese110<25016.439334646–966Higher Gleason score
Lane et al. [9]American455<2205.5NA395316–507 (interquartile)Higher Gleason score
Teloken et al. [8]Brazil64<27035.9362NANAHigher positive surgical margin rate
Salonia et al. [17]Italian673<30021.44504502–1360Higher Gleason score and advanced stage
Botto et al. [7]French431<30014.4NA44090–1580Higher Gleason score
Yamamoto et al. [18]Japanese272<30018.0NA401149–943Higher biochemical failure rate
Isom-Batz et al. [12]American326<30025.0NA385133–998Higher Gleason score
Xylinas et al. [10]French107<30019.6NANANAHigher Gleason score and advanced stage
Roder et al. [19]Danish227<31728.1NA40363–1153Higher biochemical failure rate
Imamoto et al. [11]Japanese82<38750.0NA387146–943Advanced stage

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

In the present study, we found a relationship between preoperative serum total testosterone level and RP specimen Gleason score in ethnic Chinese patients with clinically localized PCa who underwent RP. Patients with high grade disease (Gleason score ≥ 8) in RP specimens had a significantly lower preoperative total testosterone level (mean total testosterone was 320 ng/dL) and a higher rate of pathological stage T3a–T3b disease, than those with low grade disease. More importantly, the present data suggest that, in ethnic Chinese patients with localized PCa, preoperative serum total testosterone <250 ng/dL should be considered as a potential marker for more aggressive disease, because 44.4% patients with preoperative total testosterone <250 ng/dL had Gleason score ≥ 8 cancer in their prostate specimens, which was significantly higher than patients with preoperative total testosterone ≥250 ng/dL (15.2%).

Over the last 20 years, the prognostic value of pretreatment serum total testosterone level in patients with metastatic PCa has been investigated. Studies have consistently shown more aggressive disease, worse treatment response and worse prognosis in patients with metastatic PCa who had lower serum total testosterone before the beginning of androgen deprivation therapy [20–22]. More recently, some researchers showed that a lower pretreatment total testosterone level also correlated with worse clinical and pathological determinants of clinically localized PCa, including a higher Gleason score in RP specimens [7,9,10], advanced pathological stage [10–12], an increased risk of positive surgical margins [8] and a worse 5-year biochemical relapse-free survival [18,19]. It is not surprising, therefore, that a correlation between preoperative total testosterone and RP specimen Gleason score was found in the present study. The clinical value of these correlations is that pretreatment total testosterone may be used as a marker to estimate the risk of localized PCa, and help determine the management of an individual patient. For example, the ethnic Chinese patients with localized PCa with low pretreatment total testosterone (<250 ng/dL) included in the present study seem to be poor candidates for delayed RP owing to the 44.4% risk of Gleason score ≥ 8 cancer in the prostate specimen.

To make it more applicable to clinical practice, some researchers have analysed pretreatment total testosterone as a dichotomous variable and found that pretreatment serum total testosterone < 300 ng/dL (defined as hypogonadism) was an excellent predictor of advanced pathological stage, higher pathological grade and higher biochemical recurrence rate in patients with clinically localized PCa treated by RP [7,10,12,17,18]. In the present study, however, we found that setting the threshold for hypogonadism at total testosterone < 300 ng/dL was not appropriate for ethnic Chinese patients, because this threshold could not distinguish patients with more aggressive PCa from those with less aggressive disease (Table 3). This may be attributable to ethnic variability. In an endocrinological study investigating variations in testosterone levels across ethnic groups, Heald et al. [13] found that circulating serum total and free testosterone levels were significantly lower in healthy Pakistani men than in European or African-Caribbean men. The authors believed that it was necessary to establish ethnic-specific ranges for the interpretation of total and free testosterone levels in men [13]. Table 4 shows that, compared with patients from other races, our ethnic Chinese patients retain the lowest median preoperative total testosterone level (346 ng/dL). In addition, some previous studies that classified patients with pretreatment total testosterone thresholds for hypogonadism other than 300 ng/dL also reported positive results [8,9]. In the present study, therefore, we set a second threshold for hypogonadism at the level of pretreatment total testosterone <250 ng/dL and found it worked better than the first threshold (total testosterone <300 ng/dL).

In the present study, patients' preoperative total testosterone did not correlate with pathological tumour stage. Stage pT3a–T3b was noted in 44.4% patients with preoperative total testosterone < 250 ng/dL vs 35.9% patients with preoperative total testosterone ≥250 ng/dL. This may be explained by the large number of patients with advanced stage PCa in this study. In previous studies, which reported a correlation between preoperative low total testosterone and advanced pathological stage, the locally advanced cancer (stage ≥ pT3a) constituted only 21.1–29.3% of all cases [10–12]; however, in the present study, locally advanced cancer (pT3a–T3b) constituted 37.3% of all cases and the remaining patients all had organ-confined cancer (Table 1).

The reasons for the association between low serum total testosterone and more aggressive PCa are still under investigation. This association may be attributable to selection of poorly differentiated cancer cells by low androgen levels. This hypothesis comes from the Prostate Cancer Prevention Trial (PCPT) study. In the PCPT study, the 5α-reductase type 2 inhibitor finasteride increased the incidence of high grade cancer, although the overall incidence of PCa was decreased [23]. The same observation was recently made in the Reduction by Dutasteride of Prostate Cancer Events (REDUCE) trial with dutasteride (5α-reductase type 1 and type 2 inhibitor) [24]. This hypothesis has been confirmed in animal studies. Using a genetically engineered mouse model of human PCa, Banach-Petrosky et al. [25] discovered that the inhibition of 5α-reductase lowered intraprostatic levels of dihydrotestosterone, and this relatively low androgen environment induced the development of poorly differentiated PCa.

The strengths of the present study are its prospective design, its being the first report in ethnic Chinese patients, the fact that it was carried out at a single centre with no discontinuity in recruitment, and that all surgical procedures and specimen evaluations were performed using a consistent method. All blood samples were drawn after an overnight fast between 7:00 and 10:00 h, thus avoiding a potential methodological flaw because of different collection times and diurnal variation of the steroid hormones. A further strength is that we excluded patients with uncontrolled diabetes, thyroid disease, hyperprolactinaemia, hypoalbuminaemia, abnormal liver or renal function as well as those who received any kind of neoadjuvant hormonal therapy.

The present study has some limitations. One was that we did not measure free testosterone, which generally reflects the clinical situation more accurately than total testosterone levels. In addition, we did not use gas chromatography–mass spectrometry, which is considered the ‘gold standard’ for measuring circulating total testosterone levels [26]. For the specific purpose of this study and to reflect routine practice of a clinical biochemistry laboratory, we chose to measure circulating total testosterone using commercially available analytical methods. Another limitation was the limited duration of clinical follow-up for the study population, which meant we could not evaluate the biochemical recurrence rate.

In conclusion, this prospective study shows that setting the threshold for hypogonadism at the level of pretreatment serum total testosterone <300 ng/dL is not appropriate for ethnic Chinese patients with clinically localized PCa, because, in these patients, this threshold is not able to distinguish patients with more aggressive PCa from those with less aggressive disease. Setting the threshold at the level of total testosterone <250 ng/dL works better, because patients with pretreatment serum total testosterone <250 ng/dL have a higher incidence of Gleason score 8–10 disease in RP specimens. Moreover, pretreatment total testosterone should be considered as a potential prognostic factor and added to routine evaluation for ethnic Chinese patients with localized PCa.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

This study was supported in part by grants from National Natural Science Foundation of China (30801149, 30973009).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES