Chemohormonal therapy as primary treatment for metastatic prostate cancer: A randomized study of estramustine phosphate plus luteinizing hormone-releasing hormone agonist versus flutamide plus luteinizing hormone-releasing hormone agonist


Masanori Noguchi md, Department of Urology, Kurume University School of Medicine, 67 Asahi-machi, Kururume, Fukuoka 830-0011, Japan. E-mail:


Background:  The present study was undertaken mainly to investigate whether chemohormonal therapy with estramustine phosphate plus luteinizing hormone-releasing hormone (LHRH) agonist has a more beneficial effect than the hormonal therapy with flutamide plus LHRH agonist for newly diagnosed patients with metastatic prostate cancer.

Methods:  A total of 57 patients with metastatic prostate cancer aged 59–80 years (median 74 years) were entered in the study and were randomized to the treatment of estramustine phosphate (560 mg/day) plus LHRH agonist (estramustine group) or flutamide (375 mg/day) plus LHRH agonist (flutamide group) with stratification for the degree of performance status, histological differentiation and bone metastasis.

Results:  Both of the treatment regimens were well tolerated with similar incidences of adverse drug reactions. The overall response rates (complete response plus partial response) at 12 weeks after treatment in the estramustine and flutamide groups were 76 and 55%, respectively. The median time to objective progression for the estramustine group (25.4 months) was longer than that of the flutamide group (14.6 months). The serum levels of follicle stimulating hormone and testosterone were significantly lower in the estramustine group.

Conclusions:  Chemohormonal therapy with estramustine phosphate plus LHRH agonist showed longer clinical progression-free survival than the hormonal therapy with flutamide plus LHRH agonist (P = 0.03), although there was no significant difference in the overall survival. A larger-scaled trial with more statistical power is required to clarify that the former regimen is more beneficial than the latter for newly diagnosed patients with advanced prostate cancer.


Suppression of testicular androgen, either by surgical castration or by long-term use of a luteinizing hormone-releasing hormone (LHRH) agonist, is the most common treatment for newly diagnosed patients with metastatic prostate cancer. However, patients who fail the first line of endocrine therapy have a poor prognosis and their median survival is measured in months.1 Labrie et al. proposed a theory of adrenal androgen hypersensitivity rather than androgen resistance to explain the poor response, or the relapse after an initial response, by medical or surgical castration.2 Since then, numerous large clinical trials have compared testicular androgen blockade with maximum androgen blockade (MAB) by adding long-term treatment of anti-androgen, such as nilutamide, flutamide or cyproterone acetate, to testicular androgen blockade. The recent meta-analysis, which involved 98% of the worldwide randomized evidence, suggested that in advanced prostate cancer the addition of an anti-androgen will improve the absolute 5-year survival by approximately 2–3%, with a range of uncertainty of approximately 0–5%.3

On the basis of the theory that hormone-resistant cells are present in small numbers from the onset of disease, it is logical that primary treatment of prostate cancer should involve both a hormonal and a cytotoxic component, provided that chemotherapy is effective in reducing hormone-resistant cells.4 Estramustine phosphate, a combination of estrogen with nornitrogen mustard, disrupts cytoplasmic microtubules, inhibits assembly of the nuclear matrix and inhibits the multidrug-resistance transporter p-glycoprotein.5 The antimitotic effect of estramustine phosphate in vivo has also been reported.6 This antimitotic effect is believed to be more responsible for the reported anticancer effect of estramustine phosphate than its estrogenic action in patients with hormone refractory prostate cancer,4–6 but the use of estramustine phosphate as a first-line treatment for advanced prostate cancer is controversial. At the beginning of the present study, there were many discussions as to the role of MAB for the first-line treatment for advanced prostate cancer, but MAB is usually used for advanced prostate cancer in Japan; therefore, we planned the study to compare clinical efficacy and toxicity between the combination therapy of estramustine phosphate plus LHRH agonist and flutamide plus LHRH agonist for newly diagnosed patients with metastatic prostate cancer.


Patient selection

Patients with newly diagnosed and pathologically confirmed adenocarcinoma of the prostate (disease stages D1 or D2), based on the American Cancer Society classification,7 were eligible for entry into the study. Patients previously treated with orchiectomy, hormones, antitumor chemotherapy or radiotherapy were excluded. Patients were also excluded if their performance status was greater than 3 according to the Eastern Cooperative Oncology Group scale. Furthermore, patients who had brain metastasis, a life expectancy of less than 3 months, another neoplasm or severe concomitant illness (renal, hepatic, cardiovascular or neuropsychiatric disorders) were not eligible for this study.


After patients gave their informed consent, they were assigned in equal numbers to one of two treatment groups using a stratified, permuted block randomization scheme, with randomization through telephone contact. Patients were stratified according to the degree of performance status, histological differentiation and degree of bone metastasis. The two treatment groups were (i) estramustine phosphate plus LHRH agonist (either goserelin acetate or leuprolide acetate); and (ii) flutamide plus LHRH agonist. Estramustine phosphate (Estracyt; Nippon Shinyaku, Kyoto, Japan) 280 mg was administered orally b.i.d. in the morning and evening. Flutamide (Odyne; Nippon Kayaku, Tokyo, Japan) 125 mg was administered orally t.i.d. after each meal. Either goserelin acetate (Zoladex; AstraZeneca Japan, Osaka, Japan) 3.60 mg depot or leuprolide acetate (Luprin, Takeda Chemical Industries, Osaka, Japan) 3.75 mg depot was administered subcutaneously once every 4 weeks.

Evaluation of disease

Pretreatment evaluation included a baseline history and physical examination, automated complete blood count, platelet count, electrolytes, blood urea nitrogen and creatinine, biochemical profile, prostate-specific antigen (PSA), urinalysis, chest X-ray, bone scan and skeletal radiographs of abnormal areas, computed tomography (CT) scan of the pelvis, and prostate ultrasonography. PSA levels were measured by a radioimmunoassay using Hybritech Tandem-R kits (normal range 0–4.0 ng/mL). The metastatic findings on bone scans were classified into four groups in accordance with the extent of the disease (EOD) using the method described previously by Soloway et al.: grade 1, less than six bone metastases (a lesion occupying the entire vertebral body was counted as two lesions); grade 2, 6–20 bone metastases; grade 3, more than 20 bone metastases but less than a ‘superscan’ (diffuse symmetrical uptake without visualization of the kidneys); and grade 4, superscan or its equivalent (involvement of more than 75% of the ribs, vertebrae and pelvic bones).8 Patients were clinically evaluated monthly for the first 3 months and every 3 months thereafter. Biochemical tests, PSA measurements and prostate ultrasonography were re-peated with the same frequency, whereas a bone scan, chest X-ray, abdominal CT scan evaluating the liver and retroperitoneal lymph node, and the other instrumental examinations were repeated every 6 months unless specific symptoms occurred. Endocrine parameters (luteinizing hormone, follicle stimulating hormone, testosterone and estradiol) were measured at each treatment period (baseline, 4, 8 and 12 weeks).

Response and progression criteria

Objective responses were evaluated according to the modified National Prostatic Cancer Project criteria to include a decrease of PSA by ≥ 50% in the partial response (PR) category.9 A complete response (CR) was defined as normalization of the PSA level and, in patients with measurable disease, disappearance of all lesions without the occurrence of new ones. PR was defined as a decrease of ≥ 50% in the sum of the products of the longest diameters of all measurable lesions persisting for ≥ 4 weeks, improvement in bone scan findings, and reossification of lytic lesions, in addition to no increase in the size of any existing lesions and no appearance of new lesions. The PSA response required a decrease of ≥ 50% from the baseline level. Progressive disease was defined as any increase of ≥ 25% in the sum of the products of the longest diameters of any measurable lesions, or the appearance of an equivocal new lesion. Biochemical progression was considered to have occurred if PSA levels reached twice the normal level in CR patients, were elevated by ≥ 50% compared with pretreatment values in PR patients or increased by ≥ 125% of the pretreatment value in patients with no response to treatment. Stable disease was defined as no change in the size of measurable lesions and PSA (<25% deviation from the baseline value) for ≥ 1 month. When there was objective progression, the first occurrence of the event was considered for the analysis of time to objective progression. The main efficacy end-points were an objective response to treatment, time to progression, and survival. Patients continued taking the study drug until they had objective progression or intolerance, or until they withdrew consent. Adverse drug reactions were recorded at each study visit and graded using World Health Organization criteria.

Statistical analysis

The enrolment duration of the present study was 2 years with 5 years of follow up. Thirty-one patients were required in each treatment group. It was assumed that the flutamide group would have a 65% progression-free rate after 1 years. The sample size of the estramustine group provides an 80% ability to detect a 20% improvement in the progression-free rate at a one-sided 0.05 level of significance. Time to progression and survival were defined from the time of registration. Time to progression and duration of survival curves were calculated using the Kaplan–Meier technique and groups were compared using a log–rank test. Comparisons of response and safety between both the treatments groups were made using the χ2 test for the trend and Fisher's exact test, respectively. To evaluate the time-course of hormonal levels in serum, it is necessary to analyze the data at each sampling point because such levels usually change rapidly. Therefore, the data were analyzed using ancova by employing the baseline values as a covariate.


A total of 57 patients were enrolled in the trial at 19 institutions between June 1995 and March 1998. Six patients (all of them in the flutamide group) were excluded from all analyzes. The reasons for exclusion were failure of entry criteria for four patients, protocol violation for one patient (treatment with the other hormone) and refusal of treatment for one patient. The main characteristics of the patients and disease are shown in Table 1. There were no noteworthy differences between the two groups in prognostic factors including age, performance status, Gleason score, EOD grade, PSA levels and history of cardiovascular disease.

Table 1.  Distribution of patient characteristics
CharacteristicsLHRH + estramustine
n (%)
LHRH + flutamide
n (%)
  1. EOD, extent of disease; LHRH, luteinizing hormone-releasing hormone; PD, progressive disease; PR, partial response.

Not eligible 0 (0) 6 (21)
Eligible29 (100)22 (79)
Age (years)
 <70 9 (31) 5 (23)
 70–74 9 (31) 6 (27)
 >7511 (38)11 (50)
Performance status
 0–124 (83)16 (73)
 2 5 (17) 6 (27)
 3 0 (0) 0 (0)
Disease stage
 D1 1 (3) 0 (0)
 D228 (97)22 (100)
Gleason score
 2–4 1 (3) 2 (9)
 5–617 (59)12 (55)
 7–1011 (38) 8 (36)
EOD grade (bone metastasis)
 0–1 9 (31) 7 (32)
 2–319 (66)14 (64)
 4 1 (3) 1 (5)
Prostate-specific antigen (ng/mL)
 <40 5 (18) 2 (9)
 40–200 9 (31) 7 (32)
 >20015 (51)13 (59)
Clinical response
 PR22 (76)12 (55)
 Stable 6 (21) 7 (32)
 PD 1 (3) 3 (13)

Both treatment groups tolerated treatment well, which resulted in a similar incidence of adverse drug reactions between the two groups. Seven of the 29 patients in the estramustine group chose to discontinue the treatment: four for liver enzyme increase; two for cardiovascular problems; and one for diarrhea. Five of the 22 patients in the flutamide group chose to discontinue the treatment: three for liver enzyme increase and two for cardiovascular problems. Table 2 shows the adverse drug reactions experienced by patients during the study. The incidences of gastrointestinal toxicity, liver enzyme increase and cardiovascular problems were similar in the two treatment groups. The incidence of gynecomastia, however, was significantly higher in the estramustine group (P < 0.001).

Table 2.  Adverse experiences of patients categorized by treatment
Adverse experienceLHRH + estramustine (n = 29)LHRH + flutamide (n = 22)
Grade 1Grade 2Grade 3Grade 4Total (%)Grade 1Grade 2Grade 3Grade 4Total (%)
  1. LHRH, luteinizing hormone-releasing hormone.

Gastrointestinal toxicity
 Diarrhea 0010 1 (3)0000 0 (0)
 Nausea/vomiting 3200 5 (17)2000 2 (9)
Liver enzyme increase 740011 (38)910010 (45)
Cardiovascular problems 1210 4 (14)0020 2 (9)
Skin rash 0000 0 (0)0200 2 (9)
Anemia 1300 4 (14)2000 2 (9)
Gynecomastia1000010 (34)0000 0 (0)

Table 3 shows the hormonal evaluation in each treatment group at baseline, 4, 8 and 12 weeks. Serum LH, FSH and testosterone levels after treatment were significantly lower than before treatment in both groups, whereas estradiol levels after treatment were higher in the estramustine group. Moreover, the mean values of the serum testosterone levels in the estramustine group were significantly lower than in the flutamide group at 4 weeks (P = 0.04), and the FSH levels in the estramustine group were also significantly lower than those in the flutamide group at 4, 8 and 12 weeks (P < 0.0001).

Table 3.  Serum hormone levels after treatment
(ng/mL, mean + SD)
(mIU/mL, mean ± SD)
(mIU/mL, mean ± SD)
(ng/mL, mean ± SD)
  • *

    P = 0.04;

  • **

    P < 0.0001;

  • ***

    P < 0.001.

  • E, estramustine phosphate; F, flutamide; FSH, follicle stimulating hormone; LH, luteinizing hormone; LHRH, luteinizing hormone-releasing hormone.

Baseline4.73 ± 1.675.36 ± 1.676.44 ± 2.716.45 ± 3.8414.21 ± 9.0114.31 ± 7.22 24.72 ± 9.2321.99 ± 11.18
4 weeks0.09 ± 0.07*0.15 ± 0.070.57 ± 0.200.63 ± 0.24 0.44 ± 0.11** 2.35 ± 1.5645000 ± 4300***10.12 ± 0.36
8 weeks0.07 ± 0.030.09 ± 0.050.43 ± 0.110.44 ± 0.12 0.41 ± 0.03** 3.71 ± 1.8645000 ± 3200***11.56 ± 3.17
12 weeks0.07 ± 0.040.12 ± 0.080.40 ± 0.010.42 ± 0.06 0.41 ± 0.01** 5.65 ± 3.8643000 ± 2400***10.36 ± 0.91

The overall response rates (CR + PR) at 12 weeks after treatment were 76 and 55% in those treated with estramustine and flutamide, respectively. However, there were no differences between both the treatments when analyzed using the χ2 test. Median follow-up time was 26 months (range 3–56 months). Figure 1 shows the progression-free survival curves. The percentages of patients who had progression according to the protocol were 41% in the estramustine group and 73% in the flutamide group. The median time to objective progression was 25.4 months in the estramustine group and 14.6 months in the flutamide group. The progression-free survival rates were consistently higher (P = 0.03) in the estramustine group than in the flutamide group (Fig. 1). However, there were no significant differences in the overall duration of survival and time to death as a result of prostate cancer between the groups as determined by the log–rank statistical analysis (Figs 2, 3).

Figure 1.

Progression-free survival. There were 12 observed events in the estramustine group versus 16 in the flutamide group. The median time to objective progression was 25.4 months in the estramustine group versus 14.6 months in the flutamide group. (—) LHRH + Estramustine (n = 29); (---) LHRH + Flutamide (n = 22); log rank = 0.03.

Figure 2.

Overall survival. There were 14 observed deaths in the estramustine group versus 11 in the flutamide group. Median time to death was 35.9 months in the estramustine group versus 27.8 months in the flutamide group. (—) LHRH + Estramustine (n = 29); (---) LHRH + Flutamide (n = 22); log rank = 0.796.

Figure 3.

Prostate cancer survival. There were nine observed deaths as a result of prostate cancer in each treatment group. Median time to death as a result of the disease was 41.5 months in the estramustine group versus 29.8 months in the flutamide group. (—) LHRH + Estramustine (n = 29); (---) LHRH + Flutamide (n = 22); log rank = 0.41.


The most frequently encountered adverse effect in the present study was mild hepatotoxicity: 11 of the 29 cases (38%) in the estramustine group and 10 of the 22 cases (45%) in the flutamide group. Well-known adverse effects of oral estramustine phosphate therapy are nausea and vomiting, which were also observed in the present study. Cardiovascular adverse events (mild–moderate and severe) were observed in 14 and 9% of the estramustine and flutamide group, respectively. However, these cardiovascular adverse events did not cause death in the present study. Most direct comparisons of the tolerability profiles of estramustine phosphate and estrogens have shown that estramustine phosphate therapy is associated with fewer and less severe cardiovascular complications than with conventional estrogen therapy.5 In contrast to many antineoplastic agents, estramustine phosphate is rarely associated with myelosuppression.10 Therefore, estramustine phosphate may be a more appropriate treatment for patients with advanced prostate cancer.

Issacs emphasized the need for concomitant administration of chemotherapeutic agents in androgen ablation therapy and stated it could improve the survival of patients with newly diagnosed metastatic prostate cancer.11 Consideration of the nature and implications of prostate tumor cell heterogeneity of hormone or hormone-resistant tumor cells supports this concept. In the 1980s, the National Prostatic Cancer Treatment Group conducted three randomized clinical trials (Protocols 500,12 130013 and 170014) in which the role of early combined chemotherapy plus hormone therapy was studied in comparison with hormone therapy alone in the treatment of metastatic prostate cancer. Unfortunately, these studies did not demonstrate a survival benefit for patients treated with chemotherapy plus hormone therapy for prostate cancer. Moreover, second-line treatment has demonstrated that the poor sensitivity of prostate cancer to the known cytotoxic agents, as well as the toxicity of those agents, renders most established cytotoxic agents unsuitable for first-line treatment.

Recently, however, the pessimism accompanying chemotherapy in advanced prostate carcinoma is not justified by the increasing evidence of clinical efficacy.15 Increased use of the serum PSA level in monitoring drug activity in clinical trials has suggested that some older drugs previously believed to be inactive are, in fact, active in the treatment of this disease. Patients have been placed on clinical trials at an earlier stage because of the recognition that an elevation in the serum PSA level precedes clinical relapse. In 1995, when this study was planned, the criteria for determining response by the change in PSA level remained in dispute. The majority of recent studies, however, have used change in serum PSA level as a marker of progression after hormonal therapy and chemotherapy.15,16 In the present study, treatment failure was judged as an increase in PSA of at least 50% more than the nadir value or progression of the disease by standard clinical criteria.

Chemohormonal therapy with estramustine phosphate alone or in recent combination with cytotoxic drugs has shown significant success, in particular in patients with heterogeneus tumors.16 Several studies suggest estramustine phosphate has a better effect than hormonal therapy as the primary treatment for high-grade, high-stage prostate cancer.17,18 A recent study randomized 385 patients with locally advanced prostate cancer or with bone metastases to orchiectomy or orchiectomy followed by estramustine phosphate until progression.19 In that study, there was no significant difference between the two treatments in terms of time to progression or survival. However, for younger patients (aged < 73 years) with distant metastases, there was a tendency for a longer time to progression in the estramustine-treated patients than in the orchiectomy-alone patients. The present study shows a significant difference in both the treatment groups in progression-free survival; the estramustine group showed longer progression-free survival than the flutamide group (P = 0.03). Comparisons of estramustine phosphate with flutamide in 167 evaluable patients with advanced metastatic prostate cancer in three randomized trials demonstrated that estramustine phosphate was more effective than flutamide in producing objective responses in patients with advanced disease.17,20,21 In the present study, although there was no significant difference in objective response rates between the two treatment regimens, the overall response rates (CR + PR) were 76 and 55% in the estramustine and flutamide groups, respectively. These results, together with the results of previous studies, suggest that the chemohormonal therapy with estramustine phosphate plus LHRH agonist is recommended more than the combined hormone therapy with flutamide plus LHRH agonist as a treatment regimen for newly diagnosed patients with metastatic prostate cancer. This advantageous effect might be a result of the following two issues: (i) small numbers of hormone-resistant tumor cells are present in newly diagnosed metastatic prostate cancer; and (ii) the estramustine phosphate plus LHRH agonist regimen is superior to the flutamide plus LHRH agonist regime in the suppression of serum levels of FSH and testosterone as shown in the present study.

In the present small-scaled study, the patients treated with estramustine phosphate plus LHRH agonist showed a longer progression-free survival than those treated with flutamide plus LHRH agonist, although there was no significant difference between the two treatments in the overall survival. A large-scaled trial with more statistical power is needed to demonstrate that the former regimen is more beneficial than the latter for newly diagnosed patients with metastatic prostate cancer.


We acknowledge the following contributing investigators: J Miyajima, Kurume University Medical Center; J Ichiki, Yame General Hospital; S Inada, Saiseikai Fukuoka General Hospital; H Ohkuma and S Naruo, Fukuoka Prefectural Hospital in Asakura; T Motomori, Fukuoka Prefectural Hospital in Yanagawa; Y Fujii, Tagawa Hospital; J Yahara, Chikugo City Hospital; O Yoshizumi, St Maria Hospital; Y Ohoyabu, Ohomuta City General Hospital; S Ueda, Ueda Urology Clinic; T Ishimatu, Kumamoto Red Cross Hospital; Y Hamada and A Maehara, Kumamoto Central Hospital; K Yamasaki, Yamaga City Hospital; Y Kinoshita, Hitoyoshi General Hospital; Z Sakiyama, Amakusa Central Hospital; and H Hirayama and H Ogasawara, Hirayama Urology Clinic.