What androgen deprivation costs: the side of hormone therapy the survival curves do not show

Every survival gain in advanced prostate cancer is bought with castration, and the trials that decided how long castration should last are also the trials that recorded what it costs.

How long, and what the extra time buys. EORTC 22961 randomised 970 men who had already had radiotherapy and 6 months of androgen suppression to stop or continue for 2.5 more years: 5-year mortality was 19.0 percent with the short course and 15.2 percent with the long, and non-inferiority for the short course failed (hazard ratio 1.42). RTOG 92-02 found that 24 extra months improved every endpoint except overall survival, and improved that too in the Gleason 8 to 10 subgroup (45.1 against 31.9 percent at 10 years). DART 01/05 GICOR found that even with radiation doses of 76 Gy or more, 24 extra months improved 5-year overall survival from 86 to 95 percent. After surgery, RADICALS-HD found 24 months better than 6: 10-year metastasis-free survival 78.1 against 71.9 percent, hazard ratio 0.773, with grade 3 or higher toxicity 19 against 14 percent.

What the trials recorded as the cost. EORTC 22961 named fatigue, diminished sexual function and hot flushes in both arms. RADICALS-HD reported grade 3 or higher toxicity in 19 percent on the longer course. SPARTAN, in a different setting, recorded fracture in 11.7 percent on apalutamide against 6.5 percent on placebo, which is what androgen deprivation does to bone over time.

Can the cost be reduced? Three attempts.

Taking breaks. SWOG 9346 randomised 1,535 men with metastatic disease to continuous or intermittent androgen deprivation. Erectile function and mental health were better with intermittent therapy at 3 months, and not thereafter. Median survival was 5.8 years continuous against 5.1 intermittent, hazard ratio 1.10 (90 percent confidence interval 0.99 to 1.23): non-inferiority was not shown and inferiority was not excluded. In metastatic disease, continuous remains the standard.

Changing the drug. Relugolix is an oral gonadotrophin-releasing hormone antagonist. NICE TA995 recommends it within its marketing authorisation for advanced hormone-sensitive prostate cancer, alongside radiotherapy for high-risk localised or locally advanced disease, and as neoadjuvant treatment before radiotherapy, on evidence that it reduces testosterone more reliably in the long term and reduces the risk of serious cardiovascular events compared with leuprolide.

Changing the route. PATCH compared transdermal oestradiol patches with a luteinising hormone-releasing hormone agonist in 1,694 men. Castration rates were the same at 3 months (93 percent each) and faster with patches at 1 month (83 against 65 percent). Hot flushes affected 35 percent on patches against 86 percent on the agonist; gynaecomastia 86 against 38 percent. Lumbar spine bone mineral density rose 7.9 percent with patches and fell 3.0 percent with the agonist, a difference of 9.3 percentage points (95 percent confidence interval 5.3 to 13.4). Cardiovascular events, the reason oral oestrogens were abandoned, did not differ (hazard ratio 1.11, 0.80 to 1.53).

None of this is a reason to refuse hormone therapy when it is indicated. It is a reason to ask exactly how many months are being proposed and on what evidence, and to ask about bone protection, exercise and the alternatives above.