Omlin et al (2013) report the anti-tumour activity of diethylstilbestrol (DS) and abiraterone (AA) in castration-resistant prostate cancer and suggest substantial activity of AA in DS pretreated patients. Their retrospective review includes patients who were post docetaxel as well as those who were chemotherapy naive. We feel that the additional activity of AA may have been overstated. Abiraterone is always combined with a corticosteroid – and therefore the comparison is between AA and a corticosteroid vs DS and a corticosteroid. Most of the patients in the series described by Omlin et al (2013) received DS without a corticosteroid. In the paper by Ryan et al (2013) median time to progression (TTP) was doubled when AA and prednisone was compared with prednisone alone (11.1 vs 5.6 months). When the combination of dexamethasone and DS was compared to dexamethasone alone, a similar effect was seen (8.6 vs 4.5 months) in chemotherapy-naive patients (Shamash et al, 2011), see table below.

DdS* DS was added when the patient progressed.

We feel that some of the responses described could also be explained by the fact that DS was given before docetaxel and AA after. There is evidence that chemotherapy following failure of a hormone therapy allows that treatment to work again when patients are rechallenged with the same hormone therapy.

We have reviewed our data on patients who went on to receive DS and dexamethasone followed on progression by AA and prednisolone. We have identified 12 patients who had DS and dexamethasone immediately prior to AA and prednisolone. Eleven had prior treatment with docetaxel. Five out of 12 had a 50% PSA response to DS. Progression-free survival was 7.0 months (range 3.6–8.7), for the whole group it was 3.4 months (range 1.1–8.7). For subsequent AA and prednisolone only one patient responded (PFS 8.1 months). The overall PFS was 1.8 months (range 0.6–8.1). For the 11 who received docetaxel, 10 had prior DS. Five out of 10 had a 50% PSA response with a PFS of 7.0 months (range 1.0–15.0). For this group, overall PFS was 5.75 months (range 1–15). Four out of these five subsequently re-responded following docetaxel.

This supports our view that many patients who respond to hormone therapy prechemotherapy will subsequently respond to the same hormonal therapy afterwards (Shamash et al, 2008) and that cross-resistance between AA and DS when combined with steroids may be much greater than has been suggested.