Dingwei Ye, Department of Urology, Fudan University Shanghai Cancer Center, 270 Dong'an Road, Shanghai 200032, China. E-mail: email@example.com
Testosterone is the major precursor of estradiol (E2) in men. We hypothesized that, in metastatic prostate cancer, androgen deprivation therapy suppression of serum androgen to the castration level may also disrupt serum E2 level, and variation in serum E2 level might play a role in the development of castration-resistant prostate cancer. Our investigation was designed to observe the variation in circulating oestrogen and androgen levels in metastatic prostate cancer patients after combined androgen blockade, and to explore the possible clinical significance. We recruited 105 consecutive metastatic prostate cancer patients who were treated with combined androgen blockade from June to August 2011, and divided them into three groups according to different hormone-sensitivity status, including 58 hormone-sensitive prostate cancers, 27 after failure of first-line hormone therapy (androgen-independent prostate cancer) and 20 castration-resistant prostate cancers. Another 36 consecutive patients with treatment-naive metastatic prostate cancer during the same period were used as controls. Serum testosterone, E2 and E2/testosterone (E2/T) ratio were analysed and compared between the groups. After combined androgen blockade, testosterone was suppressed to a low level, regardless of different hormone sensitivity (p > 0.05). Mean serum testosterone was 4.07, 0.15, 0.11 and 0.09 ng/mL in treatment-naive, hormone-sensitive, androgen-independent and castration-resistant prostate cancer respectively. For each group, mean E2 was 33.06, 9.23, 9.13 and 15.05 pg/mL respectively. Mean E2/T was 9.58, 269.29, 292.06 and 996.67 respectively. Recovery of E2 and increased E2/T ratio were more significantly associated with combined androgen blockade failure, especially in castration-resistant prostate cancer (p < 0.001). This study indicated that metabolism of oestrogen might change during combined androgen blockade in metastatic prostate cancer patients, and oestrogen-related pathways might play a role in the development of castration-resistant prostate cancer.
The standard treatment for men who present with metastatic prostate cancer (PC) is systemic therapy. Historically, oestrogens have been widely used as a treatment of metastatic prostate cancer until their cardiovascular side effects were considered unacceptable, and luteinizing hormone-releasing hormone analogues (LHRHa) and later antagonists have been developed and become the mainstay of hormonal treatment. PC usually responds to androgen deprivation therapy (ADT), and combined androgen blockade (CAB) with surgical or medical castration plus anti-androgens appears to provide a small survival advantage when compared with ADT monotherapy (Akaza et al., 2004; Klotz et al., 2004), but on average, after 12–18 months, the malignant cells become resistant. Once castration-resistant PC (CRPC) develops, the median survival is approximately 24–36 months, and effective treatment options are limited.
Although remarkable advances have been made during the previous year in managing metastatic CRPC patients, these gains are modest. Sipuleucel-T, abiraterone, as well as docetaxel and cabazitaxel have all shown improvements in overall survival (OS) in phase III trials; however, these innovative therapies only provide a small improvement in OS (Petrylak et al., 2004; Tannock et al., 2004; De Bono et al., 2010; Kantoff et al., 2010; Parker et al., 2012; Scher et al., 2012). Research has increasingly suggested that oestrogen signalling may play a significant role in both normal and abnormal growth of the prostate gland (Ho et al., 2008; Ellem et al., 2009; Ellem & Risbridger, 2010). Correlation between circulating oestrogens and advanced PC at radical prostatectomy has been reported occasionally, indicating that serum oestrogen is a prognostic marker for tumour aggressiveness in PC (Giton et al., 2008; Salonia et al., 2011). Oestrogens and their receptors have been implicated in the development and progression of PC (Bonkhoff & Berges, 2009).
Testosterone is the major precursor of estradiol (E2) in men. Therefore, we hypothesized that, in metastatic PC, ADT suppression of serum androgen to the castration level may also disrupt serum E2 level, and variation in serum E2 level might play a role in the development of CRPC. Our investigation was designed to observe the variation in circulating oestrogen and androgen levels in metastatic PC patients after CAB, and to explore the possible clinical significance. By comparing serum testosterone, E2, as well as estradiol/testosterone (E2/T) ratio in different hormone-sensitive states, our preliminary data indicated that metabolism of oestrogen might change during CAB, whereas oestrogen and related pathways might play a role in the development of CRPC.
Materials and methods
From June to August 2011, 105 consecutive metastatic PC patients were recruited, and none had uncontrolled diabetes, thyroid disease, hyperprolactinaemia, hypoalbuminaemia or liver disease; they had received CAB for 1–97 months (mean 23.7 months). The main regimen of first-line CAB was LHRHa (namely goserelin, triptorelin or leuprorelin) plus anti-androgens (namely flutamide or bicalutamide), and failure of first-line CAB was termed androgen independent (AIPC). CRPC was defined as follows: castrate serum testosterone <0.5 ng/mL; three consecutive rises of prostate-specific antigen (PSA), 1 week apart, resulting in two 50% increases over the nadir; anti-androgen withdrawal for at least 4 weeks for flutamide and for at least 6 weeks for bicalutamide; PSA progression, despite consecutive hormonal manipulations, or progression of osseous lesions: progression or appearance of two or more lesions on bone scan or soft tissue lesions using Response Evaluation Criteria in Solid Tumours and with nodes >2 cm in diameter (Mottet et al., 2011). As suggested by the Prostate Cancer Clinical Trials Working Group (PCWG2), in situations where the bone scan findings are suggestive of a flare reaction, or apparent new lesion(s) may represent trauma, we confirm progressive disease on bone scan with a repeat scan. Confirmation is generally not necessary if multiple new areas of uptake are observed (Scher et al., 2008).
Altogether, there were 58 hormone-sensitive PCs (HSPCs), 27 AIPCs and 20 CRPCs. Another 36 consecutive patients with treatment-naive metastatic PC (TNPC) during the same period were used as controls. Age was documented for every patient, and body mass index (BMI) was also measured. Peripheral venous blood was obtained after an overnight fasting between 08:00 and 11:00 AM, and determined by radioimmunoassay, which is routinely available in our institute. Age, time, BMI, serum testosterone, as well as E2 and E2/T ratio were analysed and compared between the groups. Analyses of peripheral venous blood from another two CRPC patients receiving estramustine were also obtained. spss version 11.5 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis.
The 141 patients treated with CAB were aged 48–88 years (mean 71.3 years), and BMI ranged from 16.3 to 33.6 kg/m2 (mean 23.7 kg/m2). Age, time from onset of CAB, BMI, serum testosterone, as well as E2 and E2/T ratio are listed in Table 1. After CAB, testosterone was suppressed to a low level, regardless of different hormone sensitivity (Kruskal–Wallis test, p > 0.05). The recovery of E2 and increased E2/T ratio were of more significance with the failure of CAB, especially in CRPC (Kruskal–Wallis test, p < 0.001); however, there were no significant differences in E2 and E2/T ratio between HSPC and AIPC (p > 0.05) (Figs. 1-3). BMI gradually increased with failure of CAB, although no significance was reached. Both univariate and multivariate analyses indicated that hormone-sensitivity status had a significant correlation with variation in serum testosterone, E2 and E2/T ratio (p < 0.001). Serum E2 levels of another two CRPC patients were 4 800 and 6 215 pg/mL, respectively, with declining serum PSA level.
Table 1. General demographic characteristics and variation in serum T, E2 and E2/T ratio in different groups
Mean (95% Confidence interval)
70.7 (68.5, 73.0)
71.5 (69.4, 73.7)
72.2 (69.9, 74.5)
70.4 (66.4, 74.3)
Time since the onset of CAB (months)
14.9 (11.2, 18.6)
28.0 (18.0, 38.0)
43.4 (30.3, 56.6)
23.1 (22.2, 24.1)
23.9 (23.3, 24.6)
24.0 (22.7, 25.2)
4.07 (3.47, 4.66)
0.15 (0.11, 0.19)
0.11 (0.06, 0.15)
0.09 (0.03, 0.15)
33.06 (29.21, 36.9)
9.23 (7.5, 11)
9.13 (6.44, 11.82)
15.05 (8.02, 22.08)
9.58 (7.63, 11.51)
269.29 (150.44, 388.13)
292.06 (119.28, 464.84)
996.67 (264.59, 1728.76)
The role of oestrogens is not new, historically, oestrogens have been widely used as a treatment of metastatic prostate cancer until their cardiovascular side effects were considered unacceptable and even recently an oestrogen patch avoiding the potential cardiovascular side effects has been studied (Meyer et al., 2011). In vitro oestrogens can activate mutant androgen receptors isolated from AIPC, whereas high-dose oestrogens have achieved objective salvage responses. A recent study found E2 suppressed PC growth in mouse models through suppression of T and dihydrotestosterone; however, it is also important to note that in this study, serum levels from animals treated with 17β-estradiol alone were 636.9 ± 321.6 pg/mL (Montgomery et al., 2010), as was detected in our study, in humans, recovered, but still limited level of circulating E2 was related to the development of CRPC, but high levels of E2 achieved by estramustine decreased serum PSA. This may have been caused by the mitotic arrest of direct cytotoxic effects on the cells, perhaps through an apoptotic mechanism (Ferro et al., 1989; Robertson et al., 1996).
In our pilot study, both univariate and multivariate analyses were used to determine the relationship between age, time from onset of CAB, BMI, hormone sensitivity and variation in circulating hormones. However, only different hormone sensitivity had a significant correlation with variation in serum testosterone, E2 and E2/T ratio. Serum testosterone was suppressed to the castration level after CAB, regardless of different hormone-sensitivity status, whereas the level of E2 reached a nadir when sensitive to first-line hormone therapy, and gradually rebounded after failure of CAB. Although non-steroidal anti-androgens have been shown to increase serum testosterone, which is then aromatized peripherally to estradiol (Decensi et al., 1993), combination with LHRHa may mechanically counteract the consequences of the androgen negative feedback interruption.
More intriguingly, E2/T ratio kept increasing with the application of CAB, especially after resistance to castration, indicating that the conversion of testosterone into E2 might be more efficient with the failure of CAB. On the other hand, we found no significant changes in E2 and E2/T ratio after failure of first-line CAB. Our findings suggest that some relationship between hormone sensitivity and castration resistance remained. Various terms have been used to describe PC that relapses after initial hormonal ablation therapy, including hormone-resistant, androgen-independent and hormone-independent cancer (Isaacs & Coffey, 1981). However, the term AIPC is gradually being discarded in the era of CRPC, which is clearly characterized and widely recognized. Here, we used the old term AIPC to describe such association with hormone sensitivity, so that one can assume there is some transition phase during the development of CRPC. Progression from HSPC to AIPC may represent a step-by-step change to castration-insensitive status, and development from AIPC to CRPC means a cut point of a fundamental change. Of course, the existent of AIPC as a transition phase between HSPC and CRPC requires further discussion, but we have already found some evidence for it; if proved, this phase may allow time to prepare for CRPC, or even prevent it.
Given the fact that P450 aromatase enzyme mediates the conversion from T to E2 (Ellem et al., 2009; Ellem & Risbridger, 2010; Salonia et al., 2011), it is assumed that the changing metabolism of oestrogen during CAB could be attributed to variation in aromatase activity. Aromatase is active in adipose tissue, adrenal glands, testes and prostate (Salonia et al., 2011), which means that it acts as a potential key regulator of the ratio of androgens to oestrogens inside and outside the prostate (Ellem & Risbridger, 2010; Salonia et al., 2011). As far as we are aware, this study is the first to investigate oestrogen metabolism at the systemic level during CAB. In the present study, BMI gradually increased, although not significantly, which suggests that patients gained weight with failure of CAB. In a retrospective review in the SEARCH database of 287 men treated with radical prostatectomy between 1988 and 2009, higher BMI was associated with a trend for greater risk of progression to CRPC in multivariate analysis (Keto et al., 2011). Although our data did not provide enough information to answer the question whether BMI was related to castration resistance or just increased with time on androgen blockade, the consensus is that ADT is associated with a high prevalence of metabolic syndrome, so that variations in metabolism, obesity, aromatase activity and serum oestrogen level may be closely related to the development of CRPC.
Oestrogen-related pathways should also be taken into consideration when dealing with variation in circulating oestrogen and development of CRPC. In recent years, the role of oestrogens and their receptors in the development and progression of PC have received more interest (Nelles et al., 2011; Salonia et al., 2011). The human prostate has a dual system of oestrogen receptors (ERα and ERβ) that undergoes profound remodelling during progression of PC (Bonkhoff et al., 1999; Leav et al., 2001; Fixemer et al., 2003). A substantial loss of ERβ is encountered in CRPC, and markedly reduced levels are found in about 40% of cases. In 10% of these tumours, ERβ is undetectable (Fixemer et al., 2003). The presence of ERα in PC is a late event in disease progression (Bonkhoff et al., 1999). Increased expression of stromal ERα has been observed in pathological specimens from PC patients after ADT (Kruithof-Dekker et al., 1996), and the most significant ERα gene expression on mRNA and protein levels has been observed in metastatic lesions and CRPC (Bonkhoff et al., 1999). Semiquantitative reverse transcriptase polymerase chain reaction has shown that ERα and ERβ transcripts are differentially expressed in human PC cell lines, including LNCaP (ERα−/ERβ+), PC-3 (ERα+/ERβ+), PC3M (ERα+/ERβ+) and DU-145 (ERα−/ERβ+) (Kim et al., 2009). As it seems the expression of ERα is associated with castration-resistant tumours, while ERβ acts adversely, the varied proportion of ERα/ERβ in PC cells and elevating level of circulating E2 may collaboratively contribute to the castration-resistant growth of PC.
Methodologically, for the specific purpose of our study, to reflect the common practice of a clinical biochemistry laboratory, we elected to measure circulating sex steroids using commercially available analytical methods. A single fasting morning venous blood sample makes the assessment of the hormone milieu more user friendly in everyday clinical practice. A single assessment of circulating sex steroid concentrations may not be completely reliable clinically. However, all our tests were carried out over a period of 2 months, using consistent methods, and all blood samples were drawn correctly after an overnight fast between 08:00 and 11:00 h, thus avoiding the potential methodological flaw caused by different collection times (Wang et al., 2009). We rigorously excluded men who had uncontrolled diabetes, thyroid disease, hyperprolactinaemia, hypoalbuminaemia or liver diseases, which may act as important confounding factors when dealing with biochemical parameters.
There were some limitations to our pilot study. The study was cross-sectional and neither longitudinal nor prospective. Although no relationship was found between the duration of CAB and variations in serum testosterone and E2, a more informative design would be to measure testosterone and E2 over time to establish what happens to their levels and E2/T ratio. Furthermore, the confounder of BMI and the peripheral conversion of androgens into oestrogens were not entirely resolved. We have reported results from an exploratory analysis, and external validation and further investigation are required to determine whether the results were coincidental or causative. As a result of the observational nature, underlying mechanisms were not further examined, and the lack of such an elegant assessment may be considered a methodological flaw.
This study does not go as far as suggesting that selective ER modulators (SERMs) or aromatase inhibitors should be used to treat CRPC. Plasma E2 levels increase in CRPC, therefore, should the E2 biosynthesis pathway be a target for future treatment, or should more research be focused on blocking the enzymes that lead to E2 production? This study opens the way for us to speculate on that. However, recent studies that have looked at both aromatase inhibitors and SERMs in controlling advanced prostate cancer have not been promising in this area. To the best of our knowledge, this is the first study to indicate that metabolism of oestrogen might change during CAB at the systemic level, and oestrogen and related pathways might play a role in the development of CRPC.
This research is under the grants of, and financed by, Shanghai Outstanding Leader program in academic discipline, no. 09XD1401200; National Science Foundation of China, no. 30973009; and Shanghai medical leaders’ foundation. Special thanks should be given to our colleagues for their technical help. They are Yijun Shen, Yao Zhu, YiPing Zhu, GuoHai Shi, ChunGuang Ma, WenJun Xiao and GuoWen Lin.
The authors declare that they have no conflict of interests.
XJQ and HLZ collected the medical history of patients, carried out the analyses and drafted the manuscript. XDY, SLZ and BD participated in the design of the study and performed analyses. DWY conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.