Testosterone has been a key therapeutic target in prostate cancer for more than 60 years, and modern advances have enriched our understanding of the pathophysiological importance of androgen signalling in this disease.
It is now clear that the androgen axis plays a central role in supporting tumour growth throughout its natural history – even in so-called ‘castration-resistant’ disease, where genetic changes may permit androgen signalling to persist despite very low circulating androgen levels.
It is therefore somewhat surprising that testosterone measurement has no widely accepted place in the management of patients with prostate cancer. With this in mind, we reviewed the evidence for several potential clinical applications of this assay with the goal of making practical recommendations for use in day-to-day practice.( )
We concluded that testosterone measurement is particularly valuable in those patients receiving androgen deprivation therapy (ADT), which is the treatment mainstay for advanced prostate cancer.
Although the importance of attaining castrate serum testosterone levels during ADT is generally appreciated, few clinicians are currently utilising testosterone assays to verify optimal therapeutic response.( ) Nonetheless, we highlight the prevalence of inadequate testosterone suppression in clinical practice (up to 12.5% and 37.5% do not achieve the 50ng/dl and 20ng/dl thresholds, respectively),( ) which has been linked to more rapid disease progression.( )
Since early breakthrough testosterone elevations appear to predict subsequent testosterone rises,( ) we recommended regular testosterone measurement during initiation of ADT (e.g. every 3 months), which may be reduced in frequency or ceased for patients consistently at target for the first three readings.
Continued monitoring is advisable for those with breakthrough elevations, with a change in treatment indicated for those patients whose testosterone remains above target in the context of rising prostate-specific antigen (PSA).
We also reiterated the European Association of Urology guideline that castrate levels of serum testosterone must absolutely be demonstrated if castration-independent status is suspected.( ) Such testing is essential to exclude other reasons for lack of efficacy.
In addition, we suggested that testosterone measurement may be utilised to inform the timing of intermittent ADT. Although PSA levels are the usual trigger for reinitiation of therapy, testosterone determination could permit standardisation of the duration of re-exposure to endogenous androgens during the off-treatment interval.
Despite an emerging body of evidence supporting intermittent administration of ADT, there are still outstanding questions regarding its clinical application, and the use of testosterone level as a criterion for re-starting therapy should be explored in further studies.
We also examined several possible roles for testosterone determination outside the setting of ADT. Although elevated endogenous androgen levels have been suggested to predispose to prostate cancer, we noted that a large pooled analysis including 18 prospective studies found no significant association between prostate cancer risk and the levels of testosterone or any other sex hormone.( ) Retrospective data in a more contemporary patient population support a similar conclusion.( ) Testosterone measurement does not therefore have a place in prostate cancer risk determination or screening programmes.
Compared with PSA concentration, the PSA:testosterone ratio has been proposed to be a more specific parameter for determining prostate cancer risk. However, a recent evaluation in 439 men with normal digital rectal exam and PSA levels 4.1-20.0ng/ml that used modern biopsy protocols (≥10 biopsy cores per patient) found no association between PSA:testosterone ratio and biopsy result.( ) Therefore, we did not recommend that testosterone measurement be used as an adjunctive diagnostic test.
We similarly found insufficient evidence to recommend serum testosterone measurement when assessing the aggressiveness of prostate cancer. Although low androgen levels have been variously associated with higher Gleason score,( ) extraprostatic disease (T3-T4),( ) positive surgical margins after radical prostatectomy,( ) post-surgical recurrence ( ) and poor response to ADT,( ) other studies found no significant association with poor prognostic factors or outcome.(8)
It is our hope that these recommendations will guide the use of testosterone measurement in current practice and highlight issues warranting further investigation.
UroToday