NATURE REVIEWS

UROLOGY

VOLUME 11

SEPTEMBER 2014

527

in a double-blinded placebo-controlled

fashion to receive 150 mg of testoster-

one enanthate or placebo injections every

2 weeks for 6 months, and testosterone and

dihydrotestosterone (DHT) levels in serum

and prostatic tissue (via prostate biopsy)

before and after therapy were assessed.

the testosterone enanthate cohort, serum tes-

tosterone levels increased from a median of

282 ng/dl at baseline to 640 ng/dl at the end

of 6 month course of treatment, whereas

no marked change in serum testosterone

levels were observed in the placebo-treated

patients (282 ng/dl at baseline; 273 ng/dl

after 6 months).

On the other hand, exami-

nation of the prostate tissue demonstrated

that median tissue levels of testosterone and

DHT did not substantially increase after

TST (0.91 ng/g and 6.79 ng/g at baseline,

respectively, versus 1.55 ng/g and 6.82 ng/g

at 6 months, respectively) and were not

statistically different than the levels repor-

ted for the placebo group (2.00 ng/g and

8.15 ng/g at baseline, respectively, compared

with 0.88 ng/g and 5.10 ng/g at 6 months,

respectively).

Although serum PSA levels

increased in the TST cohort, no differences

in high-grade prostatic intraepithelial neo-

plasia (HGPIN) or prostate cancer in the

follow-up biopsy tissues were observed

between the two groups.

The authors

concluded that TST increases serum tes-

tosterone levels, but has a limited effect on

prostate-tissue androgen levels.

The research by Marks and colleagues,

together with publications by Morgentaler

and other authors,

gave clinical relevance

to the ‘prostate-saturation theory’. Insight

into the molecular basis for the prostate-

saturation theory comes from studies that

have shown that DHT becomes maximally

bound to the androgen receptor (AR) at

serum testosterone levels moderately lower

than the normal range.

7,9,10

Although none of

these studies have clearly defined a univer-

sal serum testosterone level at which pros-

tate ARs become saturated, the saturation

point is thought to be in the 150–200 ng/dl

range.

7,9,10

Identifying a universal saturation

point for all men might be difficult given

the complexity of the AR, with its various

co-activators and co-repressors, as well as

the intricacies of the interaction between

total and free serum testosterone levels.

Nevertheless, in men with serum testoster-

one levels greater than this saturation point,

increasing serum levels of this hormone

with TST should lead to a limited change

in intraprostatic testosterone or DHT levels.

Indeed, Bhasin

et al.

administered placebo

injections or supraphysiologic doses of tes-

tosterone enanthate (600 mg per week) to

43 athletes aged 20–37 years with baseline

mean serum testosterone levels from 431 to

557 ng/dl. The authors found that, although

serum testosterone levels reached a mean

of 3,244 ng/dl in the testosterone enanthate

cohort, no relevant changes in PSA levels

occurred between the groups.

Hypogonadism and cancer risk

From an epidemiological perspective,

prostate-cancer incidence increases with

age, as does the incidence of male hypo-

gonadism. Thus, the population of men

with the highest incidence of prostate

cancer is the same as the population of men

with the highest incidence of hypogona-

dism. This raises the question of whether

or not hypogonadism is a risk factor for

prostate cancer.

The observations of Huggins and Hodges

implied that prostate-cancer markers were

increased in the presence of higher levels

of testosterone; thus, one must ask if low

testosterone levels are protective against

prostate cancer. However, a review of the lit-

erature suggests the contrary: that low levels

of serum testosterone seem to present an

increased risk of prostate cancer diagnosis

and are associated with greater aggressive-

ness of disease. For example, the incidence

of biopsy-tissue-detected prostate cancer

reported in a series of 77 hypogonadal men

with a normal prostate anatomy accord-

ing to digital rectal examination (DRE)

and serum PSA levels <4.0 ng/ml was 14%,

which is higher than the rate expected in

a random population.

12–15

Furthermore,

in an examination of 117 patients with

prostate cancer detected using transrectal

ultrasonography- guided prostate biopsy,

Hoffman

et al.

demonstrated that high-

grade prostate cancer (Gleason 8 or greater)

was more likely to be found in men with

serum testosterone levels <300 ng/dl than in

patients with testosterone levels above this

threshold.

Similarly, in a series of 156 men

with newly diagnosed prostate cancer,

serum testosterone <300 ng/dl was found

to correlate with a worse Gleason score.

Likewise, postoperative pathological

status seems to correlate with serum tes-

tosterone levels. For example, in a series

with 326 patients undergoing radical

prostatectomy, Isom-Batz

et al.

demon-

strated a significant correlation between

lower preoperative testosterone levels and

advanced pathological stage (

0.01).

Similarly, Mearini and colleagues

found

that, in a series with 103 patients diagnosed

with prostate cancer, lower preoperative

serum testosterone levels correlated with

worse pathological stage following pros-

tatectomy. In addition, Massengill and

co-workers

showed that lower testoster-

one levels were an independent predictor

of extracapsular extension in a series of

879 men who underwent radical prosta-

tectomy. Another case series of 64 patients

with localized prostate cancer demon-

strated that patients with low total serum

testosterone (<270 ng/dl) were more likely

to have positive surgical margins in radical

prostatectomy specimens.

Low serum testosterone also seems to

influence treatment outcomes in men with

prostate cancer. Yamamoto

et al.

studied

272 patients with localized prostate cancer

treated with radical prostatectomy. The

investigators obtained preoperative serum

testosterone levels in all patients and

found that independent predictors of post-

operative PSA-based biochemical recur-

rence included Gleason score (

0.006),

surgical margin status (

0.0001), pre-

operative PSA level (

0.0001), and

preoperative testosterone level (

0.021).

In fact, men with a preoperative testoster-

one <300 ng/dl had a 67.8% 5-year PSA-

failure-free survival rate, whereas the rate

among men with preoperative testoster-

one levels greater than this threshold was

84.9% (

0.035). In a series of 60 men

undergoing radical prostatectomy, Kim

and co-workers

found that, in addi-

tion to being predictive of extracapsu-

lar extension with an odds ratio (OR) of

4.96 (95% CI 1.41–17.38;

0.012), pre-

operative serum testosterone <300 ng/dl

also predicted postoperative biochemical

recurrence (OR 3.64, 95% CI 1.66–2.43;

0.015). Furthermore, Ribeiro and

colleagues

demonstrated that a pretreat-

ment testosterone <300 ng/dl portended

poorer prognosis among 144 men diag-

nosed with metastatic prostate cancer, with

the overall survival duration reduced by

6 months in the hypogonadal group com-

pared with the cohort with testosterone

levels >300 ng/dl.

‘‘

…the population of men

with the highest incidence of

prostate cancer is the same as

the population … with the highest

incidence of hypogonadism

’’

PERSPECTIVES