Prospective randomised trial comparing diethylstilboestrol and flutamide in the treatment of hormone relapsed prostate cancer


Nick Burns-Cox mb chb md frcs, The Old Garage, The Square, Wickham, Hampshire PO17 5JW, England. Email:


Background: Patients with hormone relapsed prostate cancer (HRPC) are often treated with flutamide or diethylstilboestrol. However, which of these two options is the best treatment for HRPC remains unclear.

Methods: We carried out a prospective study to determine and compare the prostate-specific antigen (PSA) response and survival in patients with hormone relapsed prostate cancer (HRPC), all of whom had previously shown a good response to medical or surgical castration. The patients were randomised to treatment with diethylstilboestrol (DES) and aspirin, or the antiandrogen flutamide. In addition, quality of life was determined by interview and questionnaire.

Results: Twenty-eight patients were randomised for treatment options.There was a significantly greater fall in the PSA (65% vs 35%; P = 0.034) after treatment with diethylstilboestrol compared to treatment with flutamide. Median survival also rose after treatment with diethylstilboestrol (18 months) compared to flutamide (11 months), but this difference did not reach statistical significance. There was no difference in the quality of life parameters between the two groups. There were no cardiovascular complications in the stilboestrol group.

Conclusions: In HRPC, treatment with stilboestrol is associated with a greater PSA fall and an increase in median survival when compared to flutamide treatment.


The optimal treatment for hormone relapsed prostate cancer (HRPC) remains unclear. Options include an antiandrogen, bisphosphonates, external beam radiation, steroids, cytotoxic agents and diethylstilboestrol (DES).

Antiandrogens such as flutamide are often used as second line therapy in HRPC to block adrenal androgens. Diethylstilboestrol is an attractive alternative, because evidence shows that it may not act just by androgen ablation but also in an androgen independent manner. Firstly, DES has been reported as being effective in the treatment of carcinoma of the prostate (CaP) without testosterone levels reaching castration levels.1 Secondly, DES has been shown to be cytotoxic in vitro2 and have an antiproliferative effect on prostate cancer cell line LnCaP probably through up-regulation of cell–cell adhesion molecules3 and increased levels of TGF-β.4

Diethylstilboestrol is as effective as androgen ablation5 and more effective than flutamide alone6,7 as a primary treatment for CaP. There is, however, a dose-dependent, increasing risk of cardiovascular complications associated with DES therapy.5 The second Veterans Administration Cooperative Urological Research Group's study reported that at doses of 1 mg the cardiovascular risk falls to the same as that of a placebo1,5 with the same benefit in cancer survival as a dose of DES 5 mg daily. This risk may be further reduced by the addition of aspirin.4

There are few studies determining the effectiveness of treatment of HRPC. Flutamide subjective response rates of 15– 30% have been reported8 and a biochemical response (50% PSA drop) has been observed in 23% of patients.9 A recent randomised study comparing flutamide and prednisone showed no difference in the time to progression, overall survival and biochemical response.9 One non-randomised single arm study used a combination of low-dose stilboestrol (1 mg daily) with hydrocortisone (40 mg daily) in consecutive patients with HRPC. Eighty-six percent of patients showed a 50% or greater reduction in PSA and 83% reported an improvement in symptoms. With these results in mind, it is surprising that a randomised trial with DES in patients with HRPC has not been conducted previously.

At a dose of 1 mg DES daily, however, only 27% of patients will reach castration levels of testosterone1 and a dose of 3 mg daily is required to ensure castration levels.10 For this reason, 3 mg daily with 75 mg of aspirin was chosen as the dose for this study.

The treatment of patients with HRPC is entirely palliative and therefore aims to balance quality and quantity of life. Therefore, any study determining the effectiveness of treatment in HRPC must include some measures of quality of life.

In this prospective randomised study, we compared biochemical response, changes in quality of life and survival in patients with HRPC treated with DES 1 mg tds and 75 mg aspirin daily or flutamide 250 mg tds.


Patients were recruited who showed carcinoma of the prostate with evidence of relapse after successful primary androgen withdrawal therapy either by orchidectomy or leutinizing hormone replacement hormone analog. The definition of hormone relapse was decided by the referring clinician and was based on a rising PSA level or clinical symptoms.

After providing their informed consent, the patients were randomised to treatment with stilboestrol 1 mg tds and 75 mg of aspirin or flutamide 250 mg tds. If they had been previously treated with an LH-RH analog, this treatment was continued. Exclusion criteria included previous treatment with an antiandrogen or a history of thrombo-embolism.

Every 6  weeks each patient underwent a review and a PSA check, and answered a quality of life questionnaire focusing on pain, wellbeing and nausea.


A total of 28 patients were randomised (DES = 16; flutamide = 12).The two groups had similar characteristics (Table 1).

Table 1.  Characteristics of patients placed in randomised treatment groups (diethylstilboestrol [DES] orflutamide) in the treatment of hormone relapsed prostate cancer (HRPC)
(n = 16)
(n = 12)
  1. DES, diethylstilboestrol; PSA, prostate-specific antigen.

Follow-up (months)  18.3   18.4
Age (years)  72  76
PSA fall after androgen ablation  96.6%  92.4%
Time to primary relapse (months)  49.3   49.5
PSA at relapse261244

After second line therapy was started, a significantly greater fall in PSA in the DES group was seen compared to the flutamide group (Fig. 1).

Figure 1.

  Percentage fall in prostate specific antigen (PSA) after commencing second line treatment (P = 0.034).DES, diethylstilboestrol.

Of the 28 randomised patients, 4 were still alive (3in the DES group; 1 in the flutamide group). Interestingly, the median survival in the DES group was 18 months compared to 11 months in the flutamide group. There was, however, no statistically significant difference in survival after primary relapse between the two groups (log–rank P = 0.306) and the Kaplan–Meier survival curves are shown in Fig. 2.

Figure 2.

Survival plot after commencing second line treatment. (O) Diethylstilboestrolgroup; (□) flutamide group.

With regard to quality of life, there was no significant difference in median scores for pain, nausea and wellbeing.


This is the first randomised study comparing DES and antiandrogen therapy in the treatment of carcinoma of the prostate in patients who have relapsed after successful primary hormone deprivation therapy.

It is accepted that an association exists between the magnitude and duration of the PSA fall after primary hormone deprivation therapy and improved patient survival and outcome.11,12 This also applies to patients with metastatic prostate cancer.13 It is interesting, therefore, that we have demonstrated a significantly greater fall in the mean PSA value in the DES group compared to those randomised to flutamide. This seems to be supported by an increase in median survival of 18 months in the DES group compared to 11 months in the flutamide group. The reason the difference did not reach significance could have been the relatively small number of patients.

Estrogen therapy at doses of 5 mg or more daily is associated with an increase in mortality due to thrombo-embolic events.5 No such increase in thrombo-embolic events is seen in patients treated with estrogens of 1 mg or less.5 In the present study, a daily dose of 3 mg was used because is not known to be associated with increased rates of thrombo-embolism, but would ensure castration levels of testosterone.10 We found no evidence of increased cardiovascular complications in the DES group.

With regard to quality of life, there was no significant difference in recorded scores for wellbeing or pain. Interestingly, although there was no significant difference in the nausea score, 3 patients treated with flutamide were unable to tolerate the gastrointestinal side-effects and were excluded from the analysis due to non-compliance.

This study shows that in the treatment of HRPC diethylstilboestrol is at least as effective as the antiandrogen flutamide. There is also a significantly greater PSA fall in response to DES therapy than with flutamide. A greater PSA decrease would be expected to correspond with improved survival. Indeed, we found a medial survival of 18 months in the DES group compared with 11 months in the flutamide group. This did not, however, reach statistical significance possibly because of the lack of statistical power in the study. A multi-institutional study is required to obtain the number of patients required to determine if there is a survival advantage to patients with HRPC treated with DES and aspirin.