A phase II study of nilutamide in men with prostate cancer after the failure of flutamide or bicalutamide therapy


Nancy B. Davis, Division of Neoplastic Diseases and Related Disorders, Medical College of Wisconsin, WI, USA.
e-mail: ndavis@mcw.edu



To determine the prostate-specific antigen (PSA) response and time to PSA or radiographic progression in men with prostate cancer refractory to bicalutamide and/or flutamide therapy.


Men with histologically confirmed prostate cancer not amenable to curative surgery or radiation therapy were eligible for the study if they had radiographic or PSA progression on at least one antiandrogen (not nilutamide) despite continued androgen suppression and standard antiandrogen withdrawal periods. All men received nilutamide 150 mg/day orally for ≥ 8 weeks unless there was unacceptable toxicity or disease progression. All men were evaluated for response, safety and toxicity. Baseline PSA levels, chest X-ray, bone scan and abdominopelvic computed tomography studies were obtained; the re-evaluation included PSA levels every 4 weeks and repeated imaging every 8 weeks in those with baseline abnormalities. The chest X-ray was repeated if there were pulmonary symptoms. Nineteen men were consented and 16 were evaluable for response.


The median (range) Gleason score was 7 (6–9) and the median number of previous second-line therapies was 2 (1–4). Bicalutamide therapy had failed in all patients. At baseline, 13 (of 16 men) had radiographically evident disease, nine with diffuse osseous and four with radiographically measurable metastases. There was no grade 3/4 toxicity; the commonest grade 1/2 toxicities were constipation (three), sensory neuropathy (four), fatigue (six), and visual changes (two) involving transiently altered colour vision and sensitivity to light, respectively. Responses included three partial and 13 with progressive disease.


The study was discontinued after a planned interim analysis because nilutamide had no apparent activity. Although well tolerated, nilutamide offers benefit to few men with prostate cancer in whom bicalutamide has failed.


Response Evaluation Criteria in Solid Tumors.


Prostate cancer is the most common malignancy in men and is expected to lead to 30 350 deaths in the USA alone in 2005 [1]. While metastatic prostate cancer can be treated successfully with hormone ablation in most men, the median duration of response is only 2 years [2]. After the development of hormone-refractory prostate cancer the median survival is 16–18 months from the start of docetaxel-based chemotherapy [3,4]. Interestingly, Small et al.[5] also reported a median survival of 16 months after initiating second-line hormonal therapy with ketoconazole, although that was before docetaxel was available [6]. While chemotherapy improves the quality of life and overall survival in men with hormone-refractory prostate cancer [7,8] the role of second-line hormonal therapy remains unclear.

Antiandrogens prevent the activation of many androgen-responsive genes by competitively binding to androgen receptors in the cytoplasm of target cells, and may promote androgen-receptor expression [9]. Both steroidal and nonsteroidal antiandrogens are clinically available for use. Three oral nonsteroidal antiandrogens, flutamide, bicalutamide and nilutamide, are currently available in the USA. The benefit of these agents when used concurrently with LHRH analogues is minimal [10]. Despite the widespread use of steroidal antiandrogens combined with LHRH analogues for combined androgen blockade [10] the use of antiandrogens as a component of combined androgen therapy has waned because of the cost and toxicity of the steroidal antiandrogens [11]. However, the use of antiandrogens as initial and secondary monotherapy in minimal disease states [12] is not infrequent, as they can inhibit the binding of dihydrotestosterone and testosterone by prostate cancer cells [13], resulting in objective remission of disease.

While flutamide has a relatively short half-life necessitating frequent oral dosing, both nilutamide and bicalutamide are administered once daily. None of the agents cause a decrease in LH production, thus serum testosterone levels remain normal or slightly increased, and potency is spared when used as monotherapy. The most common side-effects of nonsteroidal antiandrogens are diarrhoea, nausea and vomiting. Rare adverse events specific to particular agents include hepatotoxicity with all, pulmonary fibrosis with bicalutamide, and decreased adaptation to darkness with nilutamide.

Responses to antiandrogen therapy after primary hormone failure, usually defined by PSA criteria, have been consistently reported [13–17]. Responses to a second or third antiandrogen have been less commonly reported; however, in two small retrospective series nilutamide resulted in both a subjective improvement and PSA responses in men progressing after previous flutamide or bicalutamide therapy [16,17].

Given such retrospective reports, we designed a prospective trial to test the hypothesis that the use of nilutamide in men with hormone-refractory prostate cancer after treatment with flutamide and/or bicalutamide had failed would induce objective remission, as shown by PSA and/or radiographic criteria.


Patients were enrolled in the study under a protocol reviewed by the institutional review boards at the University of Chicago (17 men) and Oregon Health & Science University (two men). All patients provided written informed consent. Patients were eligible if they were ≥ 18 years old, with histologically confirmed prostate cancer not amenable to curative therapy with surgery or radiation, and had evidence of progressive disease, as defined by biochemical progression, new bone scan lesions or new or progressive radiographically measurable lesions. Biochemical (PSA-only) progression was defined as sequential increases of ≥ 50% above nadir achieved on previous therapy, with the absolute PSA value being ≥ 5 ng/mL, on at least two successive measurements ≥ 2 weeks apart. Also, progression had to occur despite treatment with adequate hormonal therapy, as defined by continued treatment with an LHRH agonist or previous orchidectomy, and after treatment with either/or both flutamide or bicalutamide and the appropriate trial of antiandrogen withdrawal (4 weeks for flutamide and 6 weeks for bicalutamide). Previous radiation therapy was permitted if ≥ 4 weeks had elapsed since completing it and the criteria for progressive disease were met. Adequate organ function and an Eastern Cooperative Oncology Group performance status of 0–2 were required. Patients who had a history of previous treatment with nilutamide, any previous chemotherapy for prostate cancer, a history of parenchymal brain metastases, an inability to give signed informed consent, the presence of uncontrolled concomitant medical illness, including symptomatic congestive heart failure, unstable angina pectoris, cardiac arrhythmia, psychiatric or social conditions likely to limit compliance with study requirements, or the presence of a second active malignancy, were excluded from participation.

All patients received nilutamide (150 mg orally daily) and all treatments were given as an outpatient. A treatment cycle consisted of 4 weeks and all patients received at least two cycles (8 weeks) of treatment unless there was unacceptable toxicity or disease progression. During the first treatment cycle serum creatinine level and electrolytes were measured weekly; thereafter, measurements were every other week. Complete blood counts and liver function tests were obtained at baseline and then at the start of each treatment cycle. Patients were evaluated for toxicity at the start of each treatment cycle. Therapy was withheld for any patient experiencing grade 3 or 4 toxicity thought to be attributable to nilutamide, although no dose reductions were permitted otherwise. Therapy was resumed when the toxicity resolved to grade 2 or less; however, if resolution required >6 weeks of interrupted dosing, the patient was removed from the study. If clinical signs of hepatitis or hepatic dysfunction developed at any time, liver function tests, including transaminases and bilirubin, were conducted. Nilutamide was withheld for abnormalities of more than twice the upper limit of normal and was resumed only upon complete resolution of abnormal laboratory values, jaundice or clinical symptoms of hepatitis.

The disease was re-evaluated by PSA measurements every 4 weeks and, for those with measurable disease at baseline, by Response Evaluation Criteria in Solid Tumors (RECIST) [18] using the appropriate indicated radiographic tests every 8 weeks (i.e. after every two cycles). The PSA response was considered complete if the PSA level decreased to <0.2 ng/mL, partial if there was a 50% decrease in PSA level and progressive for a 25% increase above the nadir PSA level, with a minimum absolute increase of 5 ng/mL or back to baseline PSA level. If there was a complete or partial response, the PSA response must have been maintained for ≥ 4 weeks. The response by bone scan was considered complete if there was disappearance of all lesions, and as progression if there was evidence of one or more new lesions. All toxicity was assessed using CTC v3.0 criteria (http://ctep.cancer.gov/reporting/ctc.html).

The study was designed to evaluate the frequency and durability of PSA and objective response in men treated with nilutamide after progression of prostate cancer on either flutamide or bicalutamide antiandrogen therapy. Third-line hormonal therapies offer a response rate of 20% at best; therefore, a true response rate of ≥ 40% to a hormonal therapy would be considered significant, while a true response rate of <20% would be clinically insignificant. This study was in two stages [19], with 17 patients enrolled during the first stage. If responses were seen in four or more patients, an additional 20 patients, for a total of 37, would be accrued. Nilutamide would be considered to be active in prostate cancer progressing, after antiandrogen withdrawal of either flutamide or bicalutamide, if 11 or more of the 37 patients responded. This design afforded a 10% chance of concluding that nilutamide is an active agent if the true response rate was <20%, and a 90% chance of concluding it is active if the true response rate was 40%.


The patients’ characteristics are shown in Table 1; 19 patients were enrolled from July 2003 to April 2004, but two were found to be ineligible by PSA screening criteria and did not receive therapy as planned. Thus 17 patients were treated in the study; one was lost to follow-up before completing 8 weeks of therapy and therefore only 16 men were evaluable.

Table 1.  The patients’ characteristics (16 men)
 African-American 2
Performance Status
 0 11
 1 5
Gleason score
 6 4
 7 6
 8 4
 9 2
 Unknown 1
Disease status at diagnosis
 Metastatic 5
 Unknown 2
Sites of metastatic disease
 None 4
 Bone only 8
  Extensive 6
  Limited 2
 Bone and other sites 1
 Visceral measurable disease only 3

Patients treated had, in general, good prognostic characteristics, including a median (range) PSA level of 35.6 (4.94–151.1) ng/mL; a median Gleason score at diagnosis of 7 (6–9); a median performance status of 0; no measurable disease by RECIST criteria in 12 men; bone-only metastases in eight; and a median number of previous second-line therapies, including antiandrogen withdrawal manoeuvres but excluding palliative radiation therapy, of 2 (1–4). Second-line therapies included androgen deprivation (two patients) or initial combined androgen deprivation (one) therapy, subsequent addition of antiandrogen (nine) to LHRH agonist, antiandrogen withdrawal (six), high-dose ketoconazole (six) and other than chemotherapeutic investigational agents (one) (Table 2). All patients had previous exposure to bicalutamide and one to flutamide before bicalutamide. Also, one patient was twice treated with bicalutamide, followed by antiandrogen withdrawal.

Table 2.  Previous second-line therapies
TherapyN or value
  • *

    One patient treated with both flutamide and bicalutamide;

  • †One patient had two separate trials of antiandrogen withdrawal after treatment with bicalutamide.

Median (range) no. of therapies2 (1–4)
Antiandrogen addition*
Antiandrogen withdrawal6
High-dose ketoconazole6
Duration of response to adding antiandrogen, months
Median (range)5 (2–15)
No. of men at
<3 months1
3–6 months3
>6 months4
Antiandrogen withdrawal
No. of patients with known response6
Median (range) duration of response, months3 (2–8)
No. of patients with known response5
Median duration of response, months4 (2–10)
Other therapies, n(not chemotherapy, not diethylstilbestrol)1

In all, 13 patients completed ≥ 8 weeks (two cycles) of therapy with nilutamide, while three progressed clinically before two cycles. One patient was lost to follow-up before completing 8 weeks on therapy. The median (range) number of treatment cycles administered was 2 (1–12+). The overall response rate was three of 16 (three partial responses), all by PSA estimates. Two responses were durable, with a continued partial response after 48 and 28 weeks of follow-up, respectively. The third responder had PSA progression after 16 weeks of therapy. Thirteen patients had progression of disease, 10 of whom did so by PSA criteria alone, while three had progression by bone scan without meeting the criteria for PSA progression. One additional patient had progression by both bone scan and PSA criteria. Of the four patients with measurable disease at baseline, three progressed by PSA criteria alone and one by both PSA and bone scan. The median (range) time to progression for nonresponders was 8 (3–12) weeks (Fig. 1). By the planned statistical analysis the study was discontinued after 17 patients, as there were only three responses in the initial treatment group.

Figure 1.

The Kaplan-Meier curve depicting the time to progression.

There were no grade 3 or 4 toxicities; the most common grade 1 and 2 toxicities were neuropathy (four men), fatigue (four), visual changes (two), oedema (two) and hot flashes (two). All neuropathy was sensory; one patient was removed from the study due to paraesthesias felt to be associated with nilutamide. The remaining neuropathies were unrelated to the study drug and patients continued on the study with no further episodes. Of the two grade 1 visual toxicities, one involved altered colour perception and the other a single episode of photosensitivity. Neither was felt to be attributable to the drug. One patient had a grade 2 obstructive uropathy and two further patients had grade 2 pain; however, an evaluation showed that these were caused by disease progression. Overall, nilutamide was well tolerated.


Despite the recently reported survival benefit for chemotherapy [3,4] few treatment options are effective for more than the simple palliation of symptoms in men with hormone-refractory prostate cancer. Given the current reliance by practitioners and patients alike on PSA levels for determining the progression of disease, more men are diagnosed with hormone-refractory disease with no development of overt measurable disease or progression of bone metastases. Arbitrary labelling a patient as ‘hormone-refractory’ significantly compromises the options for further therapy. Therefore, delaying the onset of a true hormone-refractory state through the sequential use of hormonal manipulations in men with asymptomatic advanced prostate cancer is a reasonable strategy. Previous small studies and anecdotal evidence exists for the biochemical efficacy of the sequential use of antiandrogens [13–17].

We conducted the present small phase II study of nilutamide in men with progressive prostate cancer despite previous therapy with either or both flutamide and bicalutamide. This study was closed after a planned analysis showed that there were only three partial responses. While two of these responses were durable, we were unable to identify any unique clinical or pathological features to predict which patients might benefit from nilutamide as a second-line antiandrogen. In particular, neither previous antiandrogen nor previous antiandrogen withdrawal response were consistently associated with the response to sequential nilutamide therapy. Toxicity was unremarkable; specifically, there was no pulmonary toxicity in the present patients.

Overall, nilutamide was well tolerated and could continue to be considered as part of first-line combined androgen blockade therapy, in combination with an LHRH agonist. Further studies to understand the molecular basis for sequential antiandrogen responses may prove fruitful.


None declared.