Androgen replacement therapy and prostate cancer

Mario Maggi, Sexual Medicine & Andrology, University of Florence, Florence, Italy.

April 2008


Prostate cancer cells express androgen receptors that bind testosterone and, with an higher affinity, its 5a-reduced product, DHT, and that transmit proliferative signals (Crescioli et al., Endocrinology, 144:3046, 2003). The pioneering work of Huggins (Cancer Res 1:293-7, 1941) gave rise to the clinical concept of the androgen dependence of prostate cancer, with the demonstration that castration caused regression of the prostatic cancer. This finding opened a new era in the management of prostate cancer.

However, different strategies of androgen ablation for advanced prostate carcinoma (PC) are only palliative and, in fact, rarely cure patients. In addition, after few years of treatment, PC cells often acquired an androgen-independent phenotype and the majority of subjects became refractory to this therapy. During androgen deprivation, the quality of life of PC patients is substantially impaired and we now know that also the quality of their endocrine metabolism is affected. In fact, testosterone positively impacts not only on sexual life but also on bone metabolism, mood, body composition and energy level. It is now clear that androgen deprivation is comorbid with relevant pathological conditions such as diabetes mellitus or metabolic syndrome (Maggi et al., J. Sex. Med. 4:1056, 2007). In the EPIC-Norfolk prospective population study (Khaw et al., Circulation 116:2694, 2007) an inverse relationship was found between high testosterone and mortality due to cardiovascular causes and cancer, even after adjustment for confounders. A 6 nmol/L increase in testosterone level was associated with an » 14% lower risk of total mortality. Therefore clinicians should be aware that androgen ablation for prostate cancer can in a way prolong survival, but in the other one it can cause metabolic derangements, resulting in an increased morbility and overall mortality. Hence, although androgen deprivation for advanced prostate carcinoma is not even a matter of debate, the concept of treating with androgens prostate cancer survivors (i.e. cured for their PC), who are also hypogonadal, is emerging. Before analyzing the few studies available on this particular topic, some general considerations should be done.

Androgens (testosterone and its receptor superagonist DHT) are essential for prostate development during early fetal life and maintain a continuous trophic and differentiating effect during childhood and even adult life. Prostate gland, as well as other internal and external male sexual structures, is underdeveloped or even absent in clinical conditions characterized by an early-impaired androgen secretion or action, i.e. during the first trimester of gestation. A further demonstration of the androgen-dependency of the prostate is derived from the observation that hypogonadal individuals have a lower prostate volume than classmate individuals, and from the clinical data showing that blocking DHT formation through 5 a reductase inhibitors (finasteride and dutasteride) substantially decreases prostate weight. Hence, it is clear that androgens promote prostate growth. However, they also regulate prostate cell differentiation. Several clinical observations strongly indicate that the fate of a prostate cancer harboured in a low androgen milieu is, at best, not favourable: in non-organ confined PC, low T is an independent predictor of extraprostatic disease (Massengill et al., J. Urol. 169:1670, 2003); low T is often associated with an advanced pathological state (Isom-Batz et al., J. Urol., 173: 1935, 2005); in newly diagnosed PC, a decreased androgen receptor density is associated with a higher malignant potential (Schatzl et al., J. Urol, 169: 1312, 2003); hypogonadal men have an increased prevalence of PC (Morgantaler et al., JAMA, 276:1904, 1996, Morgentaler & Rhoden, Urology 68:1263, 2006); low T is associated with an increased number of biopsies positive for PC and with an increased incidence of Gleason score ³ 8 (Hoffman et al., J. Urol. 163:824, 2000). In addition, experimental studies show that the introduction of a functioning androgen receptor (AR) in an androgen insensitive PC cell lines readily induced a more benign and less invasive phenotype (Bonaccorsi et al., Steroids. 71:304, 2006). Several genes that regulate prostate cancer cell growth and invasion (as integrins, neutral endopeptidases, prostin-1, hepsin), are androgen-dependent. It is therefore possible that androgen deprivation, switching off AR modulation, induces deregulation and higher invasiveness. It is my opinion that from all these studies the final message is: a normal androgen milieu is the optimal condition for prostate differentiation and well-being. In this line is the concept of intermittent androgen suppression (IAS) for advanced PC, which was indeed proposed several years ago (Klotz et al., Cancer 587:2546, 1986). A recent Cochrane review of the available studies indicate that IAS is at least not worst than continuous androgen suppression in terms of survival, while it shows some advantage in terms of quality of life (Conti et al., Cochrane Database Syst Rev. 4: CD005009, 2007).

There is some concern among urologists about a causative role for testosterone replacement therapy (TRT) in inducing a prostate carcinoma or in unmasking a subclinical disease. However, there is no evidence that TRT causes prostate cancer. The pooled odds ratio for TRT derived from 19 randomized clinical trials was 1.09 (0.48-2.49, 95% CI) for prostate cancer and 1.19 (0.67-2.09, 95% CI) for PSA > 4 ng/dL or 1.5% increase during study (Bhasin et al., 2006 Nature Clinical Practice Endocrinology and Metabolism, 2: 146 2006). Serum PSA is an age- and androgen-dependent protein, and, therefore it is lower in hypogonadal subjects and restored to age-adjusted normal level following TRT. A further increase in PSA should trigger prostate biopsy. The more intense screening for PSA in hypogonadal subjects undergoing TRT often end up in an increased number of prostate biopsy (odds ratio 1.87). A recent study investigating in aged hypogonadal subjects the effect of TRT at the prostate level clearly indicates that testosterone substitution has little effect on prostate cell function (Marks et al., JAMA 296:2351, 2007). When TRT was compared to placebo, several androgen-sensitive genes, as PSA, AR, NKX3.1, PAP2A, VEGF, and clusterin, measured by qRT-PCR, did not change their prostatic expression  

Because there is no evidence for an association between TRT and prostate cancer, should TRT be offered to hypogonadal subjects that have been cured for prostate cancer? Although many survivors from PC with hypogonadism are not symptomatic, those who are might ask for TRT. Studies on TRT in hypogonadal individuals cured by surgery or brachytherapy for PC are dramatically limited. In fact, cumulatively, less than fifty patients have been reported up to now (Kaufman & Graydon, J Urol. 172:920, 2004; Agarwal & Oefelein, J. Urol. 173:533, 2005; Sarosdy, Cancer 109:536, 2007). However, it is important to note that no patients stopped TRT because of cancer recurrence or documented cancer progression after several years of follow-up.

In conclusions, it is my opinion that there is no evidence that TRT is contraindicated in hypogonadal individuals who does not have a prostate cancer. In addition, TRT can be cautiously considered also in some hypogonadal patients who have been successfully treated for early-stage prostate cancer.