Evaluation of Tamoxifen and Anastrozole in the Prevention of Gynecomastia and Breast Pain Induced by Bicalutamide Monotherapy of Prostate Ca
http://www.100md.com
《临床肿瘤学》
the University and National Cancer Research Institute of Genoa
University of Genoa, Genoa
University of Bari, Bari
S Anna Hospital, Como
University of Bologna, Bologna
S Maria Misericordia Hospital
University of Udine, Gemona del Friuli, Udine
City Hospital, Caltagirone
University of Cagliari, Cagliari
City Hospital, Fidenza
City Hospital, Siena
Santo Spirito Hospital, Casale Monferrato
City Hospital, Novi Ligure, Italy
ABSTRACT
PATIENTS AND METHODS: A double-blind, placebo-controlled trial was performed in patients with localized, locally advanced, or biochemically recurrent prostate cancer. Patients (N = 114) were randomly assigned to either bicalutamide (150 mg/d) plus placebo or in combination with tamoxifen (20 mg/d) or anastrozole (1 mg/d) for 48 weeks. Gynecomastia, breast pain, prostate-specific antigen (PSA), sexual functioning, and serum levels of hormones were assessed.
RESULTS: Gynecomastia developed in 73% of patients in the bicalutamide group, 10% of patients in the bicalutamide-tamoxifen group, and 51% of patients in the bicalutamide-anastrozole group (P < .001); breast pain developed in 39%, 6%, and 27% of patients, respectively (P = .006). Baseline PSA level decreased by ≥ 50% in 97%, 97%, and 83% of patients in the bicalutamide, bicalutamide-tamoxifen, and bicalutamide-anastrozole groups, respectively (P = .07); and adverse events were reported in 37%, 35%, and 69% of patients, respectively (P = .004). There were no major differences among treatments in sexual functioning parameters from baseline to month 6. Elevated testosterone levels occurred in each group; however, free testosterone levels remained unchanged in the bicalutamide-tamoxifen group because of increased sex hormone–binding globulin levels.
CONCLUSION: Anastrozole did not significantly reduce the incidence of bicalutamide-induced gynecomastia and breast pain. In contrast, tamoxifen was effective, without increasing adverse events, at least in the short-term follow-up. These data support the need for a larger study to determine any effect on mortality.
INTRODUCTION
Bicalutamide 150 mg is generally well tolerated. However, breast pain and gynecomastia occur in a significant proportion of patients and can necessitate treatment discontinuation.8 In the EPC Program, gynecomastia and breast pain occurred in 66% and 73% of patients receiving bicalutamide 150 mg, respectively, and treatment discontinuation was necessary in approximately 16% of patients.7
The mechanisms involved in the development of gynecomastia during bicalutamide 150 mg monotherapy are related to the hypergonadotropic effects of the drug.9 Increases in testosterone levels are commonly observed in men receiving antiandrogen monotherapy and are accompanied by comparable increases in the level of 17-estradiol because of aromatization of androgens in extragonadal tissues.10 Estrogens induce the benign proliferation of male breast glandular tissue, and if proliferation is long standing, irreversible hyalinization and fibrosis may occur.9 The understanding of the pathogenesis of gynecomastia during antiandrogen monotherapy provides the rationale for pharmacologic attempts to prevent glandular proliferation, either by interfering with peripheral aromatization of testosterone or by directly competing with estrogenic stimulation at tissue level.
The primary objective of this study was to investigate the efficacy of tamoxifen (Nolvadex; AstraZeneca) or anastrozole (Arimidex; AstraZeneca), a new nonsteroidal aromatase inhibitor, in preventing the development of gynecomastia and breast pain during bicalutamide 150 mg monotherapy. Secondary end points included treatment efficacy (based on PSA response rate) and safety and tolerability. The effects of combined treatment on patient quality of life (QOL) and, in particular, on sexual functioning were also studied. The evaluation of the effect of treatment on sexual functioning seemed to be opportune in view of previous findings concerning the use of tamoxifen in male breast cancer patients and reports that up to 30% of these men had developed sexual impotence.11 Pharmacodynamic evaluations were also performed. These evaluations seemed to be opportune to study major interactions among treatments and to elucidate the mechanisms of action of both tamoxifen and anastrozole.
PATIENTS AND METHODS
Treatment and Allocation Procedures
Patients received three tablets orally once a day. Possible drug combinations were bicalutamide 150 mg plus tamoxifen placebo and anastrozole placebo (bicalutamide group); bicalutamide 150 mg plus tamoxifen 20 mg and anastrozole placebo (bicalutamide-tamoxifen group); and bicalutamide 150 mg plus tamoxifen placebo and anastrozole 1 mg (bicalutamide-anastrozole group). All drugs were provided by AstraZeneca.
Treatment was continued for 48 weeks or until disease progression (or patient death), patient refusal, or the occurrence of a serious drug-related adverse event. In the absence of disease progression at week 48, patients were treated at the discretion of the local investigator. Treatment was assigned on a random basis according to a 1:1:1 ratio. Separate randomization lists were available for each center.
Evaluations
Clinical and laboratory evaluations. Physical examination and hematology and serum biochemistry (including total PSA) evaluations were performed at baseline and every 3 months thereafter.
Tumor assessments. Abdominal ultrasound, computed tomography scan, bone scan (and subsequent skeletal computed tomography scan when required), and chest x-ray were performed at baseline to exclude the presence of distant metastases or obvious pelvic disease. Radiologic assessments were repeated at the investigators' discretion whenever disease progression was suspected on the basis of PSA level.
Evaluation of gynecomastia and breast pain. Breast ultrasound or caliper were used to measure gynecomastia. Severity of gynecomastia was scored on the basis of the largest diameter as follows: grade 1, ≤ 2 cm; grade 2, more than 2 to ≤ 4 cm; grade 3, more than 4 to ≤ 6 cm; and grade 4, more than 6 cm. Evaluations were performed every 3 months. Breast pain was evaluated via direct patient questioning at each visit and was arbitrarily scored according to severity as none, mild to moderate, or severe.
Evaluation of QOL and sexual functioning. QOL was assessed using a validated,1-5 30-item questionnaire, including specific questions relative to sexual interest and functioning. Questionnaires were administered at baseline and at 3-month intervals during treatment.
Pharmacodynamic evaluations. Blood samples were taken at baseline and at 3 and 6 months. Samples were centrifuged for 10 to 15 minutes at 1,000 x g, divided into multiple aliquots, and cryopreserved at –20°C until processing. For pharmacodynamic studies, aliquots were collected at a central laboratory (RBM-LCG, Ivrea, Italy). Testosterone, free testosterone, 17-estradiol, and sex hormone–binding protein concentration were assessed through radioimmunoassays.
Statistical and Ethical Considerations
Justification of patient numbers. The incidence of gynecomastia was the primary end point of this trial. In the previous trials performed by us in comparable groups of patients treated with either goserelin or goserelin plus flutamide, the incidence of gynecomastia was approximately 20%.12 This rate represents the percentage of patients developing gynecomastia during frontline treatment with castration or maximal androgen blockade (ie, the most common therapeutic options).
Because gynecomastia was observed in more than 50% of patients treated with bicalutamide 150 mg monotherapy in one of the previous trials,4 it was determined that anastrozole or tamoxifen would be defined as effective should the combination of one of these drugs with bicalutamide decrease the expected incidence of gynecomastia by more than 50% (eg, from 50% to 20%). Under these conditions, 47 patients per arm (total of 141 patients) were required to detect such an effect with a power of 80% and a two-sided {alpha} error of 5%.
However, recruitment was stopped after the enrollment of 114 patients because breast ultrasound detected gynecomastia in a larger proportion of patients than expected, and the actual size of the trial was sufficient to detect the required difference (≥ 50% decrease). Stopping of recruitment also seemed to be opportune because of the preliminary results of a comparable study conducted by Saltzstein et al,13 which showed anastrozole to be much less effective than tamoxifen in preventing gynecomastia and breast pain.
Statistical methods. The {chi}2 test (2 df P value; 1 df P value for pairwise comparisons) and Fisher's exact test were used to compare groups in respect to the incidence of gynecomastia, breast pain, adverse events, and PSA response rate. Each experimental group was compared with the placebo group. The multivariate logistic regression model14 was used to adjust predicted PSA response for baseline PSA level and disease status at entry.
The time spent without gynecomastia was calculated with the Kaplan-Meier method,15 and curves were compared using the log-rank test16 (2 df P value; 1 df P value for pairwise comparisons). All pairwise comparisons were adjusted with the Bonferroni correction.14 In this way, each one of the three comparisons (bicalutamide v bicalutamide-tamoxifen, bicalutamide v bicalutamide-anastrozole, and bicalutamide-tamoxifen v bicalutamide-anastrozole) was performed with a nominal two-sided alpha of 1.7%.
Repeated-measures analysis of variance,17 using the SPSS multivariate analysis of variance procedure (SPSS, Inc, Chicago, IL), was used to test for arm differences in mean sexual functioning scores and hormone levels changes across time (from baseline up to the month 6 since randomization). In this case, time was the within-subject factor, whereas assigned treatment was the between-subject factor. The Pillai's trace test on the interaction term (time by treatment) was used to compare arms (4 df P value; 2 df P value for pairwise comparisons).
All P values were two sided. Only P ≤ .05 have been reported in the text and/or in Tables and Figures.
Ethical considerations. The trial was conducted in accordance with the Declaration of Helsinki (1964) as amended in Hong Kong (1989). Each patient gave written informed consent, and the protocol was independently reviewed and approved by the ethics committee of each participating center.
RESULTS
Gynecomastia and Breast Pain
All patients with at least two consecutive measurements were used in the analyses, with the exception of seven patients (five in the bicalutamide-tamoxifen group and one each in the other two groups) who were found to already have gynecomastia at the time of their first breast ultrasonography. These patients were maintained on study, but they were evaluated separately.
Figures 1 and 2 show the incidence of gynecomastia and breast pain, respectively, in each group. More patients developed gynecomastia in the bicalutamide group compared with those in either the bicalutamide-tamoxifen group or the bicalutamide-anastrozole group. However, the difference was significant only in respect to patients in the bicalutamide-tamoxifen group (P < .001; Fig 1). Similarly, breast pain was more frequent and severe in the patients in the bicalutamide group than in patients in either the bicalutamide-tamoxifen group or the bicalutamide-anastrozole group; again, the difference was statistically significant only versus the patients in the bicalutamide-tamoxifen group (P = .001; Fig 2).
Figure 3 shows the actuarial curves for the time spent without gynecomastia, independent of the method of assessment. Both patients receiving tamoxifen and those assigned to anastrozole were free from gynecomastia for a significantly longer period of time than patients assigned to bicalutamide and placebo; however, the effect of tamoxifen was significantly greater than that of anastrozole. Separate analysis according to the method used to measure gynecomastia yielded comparable results (data not shown). All seven patients diagnosed with gynecomastia at the beginning of the trial had grade 1 gynecomastia.
The patient assigned to bicalutamide and placebo had monolateral gynecomastia and, after 6 months, developed bilateral gynecomastia. Of the five patients affected by gynecomastia in the bicalutamide-tamoxifen group, three improved, and two remained stable. The patient assigned to bicalutamide-anastrozole remained stable.
PSA Response
Obviously, this analysis excluded the patients for whom only the baseline value was available (total, 12 patients; three, four, and five patients in the bicalutamide, bicalutamide-tamoxifen, and bicalutamide-anastrozole groups, respectively) and those patients with unmeasurable PSA levels (PSA < 0.2 ng/mL) at the beginning (total, seven patients; two, three, and two patients in the bicalutamide, bicalutamide-tamoxifen, and bicalutamide-anastrozole groups, respectively). Therefore, 94 patients (35, 30, and 29 patients in the bicalutamide, bicalutamide-tamoxifen, and bicalutamide-anastrozole groups, respectively) were included in the PSA analysis.
There was no major difference among groups in the proportion of patients achieving a ≥ 50% reduction in baseline PSA level, irrespective of the baseline cutoff value selected for this analysis (Table 2). However, a nonsignificant effect favored patients assigned to bicalutamide plus placebo or bicalutamide-tamoxifen compared with patients in the bicalutamide-anastrozole group when a ≥ 80% reduction in baseline PSA level was chosen for the analysis. This effect was maintained when results were adjusted by disease stage or pretreatment PSA level (data not shown).
Treatment Safety
Overall, treatments were well tolerated, with only 10%, 11%, and 14% of patients in the bicalutamide, bicalutamide-tamoxifen, and bicalutamide-anastrozole groups, respectively, experiencing serious adverse events. However, the incidence of adverse events was higher in the bicalutamide-anastrozole group when the analysis included all events, irrespective of severity (Table 3).
QOL and Sexual Functioning
This analysis was restricted to the patients who had answered the questionnaires administered after 6 months of treatment. Although there were no major differences among groups in mean sexual interest scores, there was just a borderline (P = .05) difference relative to sexual functioning scores. Indeed, scores increased in the bicalutamide plus placebo and bicalutamide-anastrozole groups, whereas they remained unchanged in the men assigned to bicalutamide-tamoxifen treatment (Table 4).
Pharmacodynamics
Again, this analysis was restricted to the patients for whom 6-month evaluations were available (Fig 4). Testosterone levels increased in all groups. However, there was no significant difference in the changes from baseline to month 6 among groups. The behavior of free testosterone in the patients assigned to bicalutamide and placebo or bicalutamide-anastrozole paralleled that of testosterone. In the patients assigned to tamoxifen, free testosterone levels remained substantially unchanged. The difference in change across time was statistically significant (P = .045). As expected, there was an increase in 17-estradiol levels in the bicalutamide and bicalutamide-tamoxifen groups, whereas there was a slight decrease in the levels of this hormone in the patients assigned to anastrozole. Overall, the difference in change across time was highly significant (P = .009).
Sex hormone–binding globulin levels sharply increased in the patients assigned to tamoxifen, whereas they decreased in patients assigned to placebo and remained substantially unchanged in the men assigned to anastrozole. Overall, the difference in change across time was highly significant (P = .006).
DISCUSSION
Tamoxifen was extremely effective in preventing the development of bicalutamide-induced gynecomastia and breast pain and was much more effective than anastrozole. These findings are in agreement with those reported by Saltzstein et al,13 whose prior study randomized 107 patients, who were receiving bicalutamide 150 mg as adjuvant to local treatment, to concurrently receive placebo, tamoxifen, or anastrozole for 3 months at the same daily doses used in the present study. At the end of 3 months of treatment, gynecomastia and breast pain had developed in 69% of patients assigned to placebo, 64% of patients assigned to anastrozole, and only 12% of patients assigned to tamoxifen, as assessed by physical examination and questionnaire.
In this study, patients on placebo and those initially assigned to tamoxifen or anastrozole who developed gynecomastia after discontinuation of medication were re-treated with tamoxifen or anastrozole for another 3 months. An objective improvement of gynecomastia was achieved in 65% and 80% of the patients initially assigned to placebo or tamoxifen and re-treated with this antiestrogen and 20% of patients re-treated with anastrozole13 (on file, AstraZeneca, 2002).
In our trial, five patients assigned to tamoxifen were found to have gynecomastia at the time of randomization. Three of these patients improved during treatment with tamoxifen. Despite small numbers, our trial findings support previous ones and confirm that tamoxifen is likely to be effective not only in the prophylaxis but also in the treatment of gynecomastia.
This trial was not specifically designed to compare the efficacy of the combinations on study with that of bicalutamide monotherapy. Indeed, this would require a much larger number of patients and a much longer follow-up time. Moreover, survival would represent a more appropriate end point. Nonetheless, it is reassuring that no major differences were recorded in PSA response rates, in particular between the patients assigned to bicalutamide and those assigned to bicalutamide-tamoxifen, irrespective of the cutoff value chosen to define response (≥ 50% v ≥ 80%) and of tumor stage and PSA baseline level.
Both tamoxifen and anastrozole have well-documented side effects that might be additive to those produced by bicalutamide 150 mg monotherapy, which would adversely affect treatment safety and tolerability. It is reassuring that few serious adverse events were reported in the present study and that they were independent of randomized treatment. Present data also suggest that bicalutamide 150 mg plus tamoxifen may have a better safety profile than bicalutamide 150 mg plus anastrozole and that the addition of tamoxifen to bicalutamide 150 mg does not increase the number and severity of adverse events.
It has been previously shown that bicalutamide 150 mg monotherapy can allow patients to preserve a better QOL. Four studies comparing bicalutamide 150 mg monotherapy with either castration or maximal androgen blockade and using the same 30-item questionnaire have shown that bicalutamide 150 mg monotherapy has benefits with respect to physical capacity, social functioning, emotional well-being, vitality, and sexual functioning.1-5
Although a detailed QOL analysis will form the object of a separate report, it can be anticipated that the addition of either tamoxifen or anastrozole to bicalutamide had no major detrimental effect on QOL (data not shown). In particular, the addition of tamoxifen did not worsen sexual interest and functioning. This is reassuring in view of a previous reporting of 29% sexual impotence in men treated with tamoxifen for breast cancer.11
The results of all pharmacodynamic assessments will be reported in full elsewhere. Here, the findings for those hormones known to be involved both in the development of gynecomastia and in the control of prostate cancer tumor growth are discussed. The present study confirms that antiandrogen monotherapy causes an increase in testosterone and free testosterone levels and that this surge is paralleled by a comparable increase in 17-estradiol levels because of the peripheral aromatization of testosterone.11 Our findings also confirm that tamoxifen and anastrozole are likely to affect the development of gynecomastia in different ways. The fact that 17-estradiol levels remained substantially unchanged in the men assigned to bicalutamide-anastrozole, whereas they increased in men assigned to bicalutamide alone, confirms that the activity of anastrozole on the breast is mainly related to its ability to interfere with 17-estradiol synthesis.
By contrast, a peripheral effect of tamoxifen at the receptor level is suggested by the high efficacy shown by this drug despite such a sharp increase in 17-estradiol levels. However, because a parallel increase of sex hormone–binding globulin occurred in this treatment group and because of the high binding affinity of this protein for circulating 17-estradiol, it cannot be ruled out that tamoxifen might work by reducing the bioavailability of estrogens as well. Unfortunately, we had not planned to dose free 17-estradiol and were not able to document an effect comparable to the effect achieved by bicalutamide-tamoxifen on free testosterone levels.
Whatever the mechanisms involved, other endocrine changes are likely to be involved in the development of gynecomastia beyond the progressive increase in 17-estradiol levels. It is noteworthy in this regard that the levels of prolactin, which is known to exert a permissive effect on estrogen activity at breast tissue level,18 were slightly increased by the administration of bicalutamide-anastrozole, whereas they were significantly lowered by the administration of bicalutamide-tamoxifen (data not shown). In view of the mechanisms involved in the antiproliferative activity of bicalutamide, maintenance of free testosterone levels in the same range of baseline values in the bicalutamide-tamoxifen group was also reassuring in respect to treatment antitumor efficacy.
In conclusion, the results of our study should be considered with caution because they are still preliminary, and overinterpretation of the findings should be avoided. However, it is clear that bicalutamide-induced gynecomastia and breast pain can be prevented by the concurrent administration of tamoxifen; anastrozole is much less effective in this regard. This beneficial effect can be achieved without altering PSA response rates and safety and without compromising sexual functioning. Pharmacodynamic studies provide reassurance on the putative hazard resulting from the early testosterone surge induced by bicalutamide 150 mg plus tamoxifen because no change in free testosterone was observed. However, safety and efficacy issues need to be investigated through more appropriate trials.
Authors' Disclosures of Potential Conflicts of Interest
Acknowledgment
We thank Simona Barozzi for her assistance in data management and secretarial support and Matteo Puntoni, BS, for cooperating in statistical analysis. We acknowledge the assistance of Vanasia (Milan, Italy) for data managing and Opera CRO (Genoa, Italy) for data monitoring. Finally, we thank AstraZeneca (Milan, Italy) for funding the trial and allowing us full autonomy and responsibility in designing the study and analyzing and interpreting the data.
NOTES
Presented in part at the 39th Annual Meeting of the American Society of Clinical Oncology, Chicago, IL, May 31-June 3, 2003.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Iversen P, Tyrrell CJ, Kaisary AV, et al: Casodex (bicalutamide) 150-mg monotherapy compared with castration in patients with previously untreated nonmetastatic prostate cancer: Results from two multicenter randomized trials at a median follow-up of 4 years. Urology 51:389-396, 1998
2. Iversen P, Tyrrell CJ, Kaisary AV, et al: Bicalutamide monotherapy compared with castration in patients with nonmetastatic locally advanced prostate cancer: 6.3 years of followup. J Urol 164:1579-1582, 2000
3. Tyrrell CJ, Kaisary AV, Iversen P, et al: A randomised comparison of ‘Casodex’ (bicalutamide) 150 mg monotherapy versus castration in the treatment of metastatic and locally advanced prostate cancer. Eur Urol 33:447-456, 1998
4. Boccardo F, Rubagotti A, Barichello M, et al: Bicalutamide monotherapy versus flutamide plus goserelin in prostate cancer patients: Results of an Italian Prostate Cancer Project study. J Clin Oncol 17:2027-2038, 1999
5. Chatelain C, Fourcade RO, Delchambre J: Bicalutamide (Casodex) versus combined androgen blockade (CAB): Open French multicentre study in patients with metastatic prostate cancer. Br J Urol 80:283, 1997 (abstr 1111)
6. See WA, McLeod D, Iversen P, et al: The bicalutamide Early Prostate Cancer Program: Demography. Urol Oncol 6:43-47, 2001
7. See WA, Wirth M, McLeod DG, et al: Bicalutamide as immediate therapy, either alone or as adjuvant to standard care of patients with localized or locally advanced prostate cancer: First analysis of the Early Prostate Cancer Program. J Urol 168:429-435, 2002
8. Tyrrell CJ: Tolerability and quality of life aspects with the anti-androgen Casodex (ICI 176,334) as monotherapy for prostate cancer: International Casodex Investigators. Eur Urol 26:15-19, 1994
9. McLeod DG, Iversen P: Gynecomastia in patients with prostate cancer: A review of treatment options. Urology 56:713-720, 2000
10. Decensi A, Torrisi R, Fontana V, et al: Long-term endocrine effects of administration of either a non-steroidal antiandrogen or a luteinizing hormone-releasing hormone agonist in men with prostate cancer. Acta Endocrinol (Copenh) 129:315-321, 1993
11. Anelli TF, Anelli A, Tran KN, et al: Tamoxifen administration is associated with a high rate of treatment-limiting symptoms in male breast cancer patients. Cancer 74:74-77, 1994
12. Boccardo F, Pace M, Rubagotti A, et al: Goserelin acetate with or without flutamide in the treatment of patients with locally advanced or metastatic prostate cancer: The Italian Prostatic Cancer Project (PONCAP) Study Group. Eur J Cancer 29A:1088-1093, 1993
13. Saltzstein D, Cantwell A, Sieber P, et al: Prophylactic tamoxifen significantly reduces the incidence of bicalutamide-induced gynecomastia and breast pain. Br J Urol 90:120-121, 2002 (abstr PD-4.07)
14. Armitage P, Berry G: Statistical Methods in Medical Research (ed 3). Oxford, United Kingdom, Blackwell, 1994
15. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
16. Peto R, Pike MC, Armitage P, et al: Design and analysis of randomized clinical trials requiring prolonged observation of each patient: II. Analysis and examples. Br J Cancer 35:1-39, 1977
17. Healy MJR: Some problems of repeated measurements, in Bithell JF, Coppi R (eds): Perspectives in Medical Statistics. London, United Kingdom, Academic Press, 1981, pp 155-171
18. Melis GB, Paoletti AM, Petacchi FD, et al: Physiological development, benign diseases and cancer of human breast: Importance of prolactin secretion, in Melis GB (ed): Postmenopausal Hormonal Therapy: Benefits and Risk. New York, NY, Raven Press, 1987, pp 55-65(F. Boccardo, A. Rubagotti)
University of Genoa, Genoa
University of Bari, Bari
S Anna Hospital, Como
University of Bologna, Bologna
S Maria Misericordia Hospital
University of Udine, Gemona del Friuli, Udine
City Hospital, Caltagirone
University of Cagliari, Cagliari
City Hospital, Fidenza
City Hospital, Siena
Santo Spirito Hospital, Casale Monferrato
City Hospital, Novi Ligure, Italy
ABSTRACT
PATIENTS AND METHODS: A double-blind, placebo-controlled trial was performed in patients with localized, locally advanced, or biochemically recurrent prostate cancer. Patients (N = 114) were randomly assigned to either bicalutamide (150 mg/d) plus placebo or in combination with tamoxifen (20 mg/d) or anastrozole (1 mg/d) for 48 weeks. Gynecomastia, breast pain, prostate-specific antigen (PSA), sexual functioning, and serum levels of hormones were assessed.
RESULTS: Gynecomastia developed in 73% of patients in the bicalutamide group, 10% of patients in the bicalutamide-tamoxifen group, and 51% of patients in the bicalutamide-anastrozole group (P < .001); breast pain developed in 39%, 6%, and 27% of patients, respectively (P = .006). Baseline PSA level decreased by ≥ 50% in 97%, 97%, and 83% of patients in the bicalutamide, bicalutamide-tamoxifen, and bicalutamide-anastrozole groups, respectively (P = .07); and adverse events were reported in 37%, 35%, and 69% of patients, respectively (P = .004). There were no major differences among treatments in sexual functioning parameters from baseline to month 6. Elevated testosterone levels occurred in each group; however, free testosterone levels remained unchanged in the bicalutamide-tamoxifen group because of increased sex hormone–binding globulin levels.
CONCLUSION: Anastrozole did not significantly reduce the incidence of bicalutamide-induced gynecomastia and breast pain. In contrast, tamoxifen was effective, without increasing adverse events, at least in the short-term follow-up. These data support the need for a larger study to determine any effect on mortality.
INTRODUCTION
Bicalutamide 150 mg is generally well tolerated. However, breast pain and gynecomastia occur in a significant proportion of patients and can necessitate treatment discontinuation.8 In the EPC Program, gynecomastia and breast pain occurred in 66% and 73% of patients receiving bicalutamide 150 mg, respectively, and treatment discontinuation was necessary in approximately 16% of patients.7
The mechanisms involved in the development of gynecomastia during bicalutamide 150 mg monotherapy are related to the hypergonadotropic effects of the drug.9 Increases in testosterone levels are commonly observed in men receiving antiandrogen monotherapy and are accompanied by comparable increases in the level of 17-estradiol because of aromatization of androgens in extragonadal tissues.10 Estrogens induce the benign proliferation of male breast glandular tissue, and if proliferation is long standing, irreversible hyalinization and fibrosis may occur.9 The understanding of the pathogenesis of gynecomastia during antiandrogen monotherapy provides the rationale for pharmacologic attempts to prevent glandular proliferation, either by interfering with peripheral aromatization of testosterone or by directly competing with estrogenic stimulation at tissue level.
The primary objective of this study was to investigate the efficacy of tamoxifen (Nolvadex; AstraZeneca) or anastrozole (Arimidex; AstraZeneca), a new nonsteroidal aromatase inhibitor, in preventing the development of gynecomastia and breast pain during bicalutamide 150 mg monotherapy. Secondary end points included treatment efficacy (based on PSA response rate) and safety and tolerability. The effects of combined treatment on patient quality of life (QOL) and, in particular, on sexual functioning were also studied. The evaluation of the effect of treatment on sexual functioning seemed to be opportune in view of previous findings concerning the use of tamoxifen in male breast cancer patients and reports that up to 30% of these men had developed sexual impotence.11 Pharmacodynamic evaluations were also performed. These evaluations seemed to be opportune to study major interactions among treatments and to elucidate the mechanisms of action of both tamoxifen and anastrozole.
PATIENTS AND METHODS
Treatment and Allocation Procedures
Patients received three tablets orally once a day. Possible drug combinations were bicalutamide 150 mg plus tamoxifen placebo and anastrozole placebo (bicalutamide group); bicalutamide 150 mg plus tamoxifen 20 mg and anastrozole placebo (bicalutamide-tamoxifen group); and bicalutamide 150 mg plus tamoxifen placebo and anastrozole 1 mg (bicalutamide-anastrozole group). All drugs were provided by AstraZeneca.
Treatment was continued for 48 weeks or until disease progression (or patient death), patient refusal, or the occurrence of a serious drug-related adverse event. In the absence of disease progression at week 48, patients were treated at the discretion of the local investigator. Treatment was assigned on a random basis according to a 1:1:1 ratio. Separate randomization lists were available for each center.
Evaluations
Clinical and laboratory evaluations. Physical examination and hematology and serum biochemistry (including total PSA) evaluations were performed at baseline and every 3 months thereafter.
Tumor assessments. Abdominal ultrasound, computed tomography scan, bone scan (and subsequent skeletal computed tomography scan when required), and chest x-ray were performed at baseline to exclude the presence of distant metastases or obvious pelvic disease. Radiologic assessments were repeated at the investigators' discretion whenever disease progression was suspected on the basis of PSA level.
Evaluation of gynecomastia and breast pain. Breast ultrasound or caliper were used to measure gynecomastia. Severity of gynecomastia was scored on the basis of the largest diameter as follows: grade 1, ≤ 2 cm; grade 2, more than 2 to ≤ 4 cm; grade 3, more than 4 to ≤ 6 cm; and grade 4, more than 6 cm. Evaluations were performed every 3 months. Breast pain was evaluated via direct patient questioning at each visit and was arbitrarily scored according to severity as none, mild to moderate, or severe.
Evaluation of QOL and sexual functioning. QOL was assessed using a validated,1-5 30-item questionnaire, including specific questions relative to sexual interest and functioning. Questionnaires were administered at baseline and at 3-month intervals during treatment.
Pharmacodynamic evaluations. Blood samples were taken at baseline and at 3 and 6 months. Samples were centrifuged for 10 to 15 minutes at 1,000 x g, divided into multiple aliquots, and cryopreserved at –20°C until processing. For pharmacodynamic studies, aliquots were collected at a central laboratory (RBM-LCG, Ivrea, Italy). Testosterone, free testosterone, 17-estradiol, and sex hormone–binding protein concentration were assessed through radioimmunoassays.
Statistical and Ethical Considerations
Justification of patient numbers. The incidence of gynecomastia was the primary end point of this trial. In the previous trials performed by us in comparable groups of patients treated with either goserelin or goserelin plus flutamide, the incidence of gynecomastia was approximately 20%.12 This rate represents the percentage of patients developing gynecomastia during frontline treatment with castration or maximal androgen blockade (ie, the most common therapeutic options).
Because gynecomastia was observed in more than 50% of patients treated with bicalutamide 150 mg monotherapy in one of the previous trials,4 it was determined that anastrozole or tamoxifen would be defined as effective should the combination of one of these drugs with bicalutamide decrease the expected incidence of gynecomastia by more than 50% (eg, from 50% to 20%). Under these conditions, 47 patients per arm (total of 141 patients) were required to detect such an effect with a power of 80% and a two-sided {alpha} error of 5%.
However, recruitment was stopped after the enrollment of 114 patients because breast ultrasound detected gynecomastia in a larger proportion of patients than expected, and the actual size of the trial was sufficient to detect the required difference (≥ 50% decrease). Stopping of recruitment also seemed to be opportune because of the preliminary results of a comparable study conducted by Saltzstein et al,13 which showed anastrozole to be much less effective than tamoxifen in preventing gynecomastia and breast pain.
Statistical methods. The {chi}2 test (2 df P value; 1 df P value for pairwise comparisons) and Fisher's exact test were used to compare groups in respect to the incidence of gynecomastia, breast pain, adverse events, and PSA response rate. Each experimental group was compared with the placebo group. The multivariate logistic regression model14 was used to adjust predicted PSA response for baseline PSA level and disease status at entry.
The time spent without gynecomastia was calculated with the Kaplan-Meier method,15 and curves were compared using the log-rank test16 (2 df P value; 1 df P value for pairwise comparisons). All pairwise comparisons were adjusted with the Bonferroni correction.14 In this way, each one of the three comparisons (bicalutamide v bicalutamide-tamoxifen, bicalutamide v bicalutamide-anastrozole, and bicalutamide-tamoxifen v bicalutamide-anastrozole) was performed with a nominal two-sided alpha of 1.7%.
Repeated-measures analysis of variance,17 using the SPSS multivariate analysis of variance procedure (SPSS, Inc, Chicago, IL), was used to test for arm differences in mean sexual functioning scores and hormone levels changes across time (from baseline up to the month 6 since randomization). In this case, time was the within-subject factor, whereas assigned treatment was the between-subject factor. The Pillai's trace test on the interaction term (time by treatment) was used to compare arms (4 df P value; 2 df P value for pairwise comparisons).
All P values were two sided. Only P ≤ .05 have been reported in the text and/or in Tables and Figures.
Ethical considerations. The trial was conducted in accordance with the Declaration of Helsinki (1964) as amended in Hong Kong (1989). Each patient gave written informed consent, and the protocol was independently reviewed and approved by the ethics committee of each participating center.
RESULTS
Gynecomastia and Breast Pain
All patients with at least two consecutive measurements were used in the analyses, with the exception of seven patients (five in the bicalutamide-tamoxifen group and one each in the other two groups) who were found to already have gynecomastia at the time of their first breast ultrasonography. These patients were maintained on study, but they were evaluated separately.
Figures 1 and 2 show the incidence of gynecomastia and breast pain, respectively, in each group. More patients developed gynecomastia in the bicalutamide group compared with those in either the bicalutamide-tamoxifen group or the bicalutamide-anastrozole group. However, the difference was significant only in respect to patients in the bicalutamide-tamoxifen group (P < .001; Fig 1). Similarly, breast pain was more frequent and severe in the patients in the bicalutamide group than in patients in either the bicalutamide-tamoxifen group or the bicalutamide-anastrozole group; again, the difference was statistically significant only versus the patients in the bicalutamide-tamoxifen group (P = .001; Fig 2).
Figure 3 shows the actuarial curves for the time spent without gynecomastia, independent of the method of assessment. Both patients receiving tamoxifen and those assigned to anastrozole were free from gynecomastia for a significantly longer period of time than patients assigned to bicalutamide and placebo; however, the effect of tamoxifen was significantly greater than that of anastrozole. Separate analysis according to the method used to measure gynecomastia yielded comparable results (data not shown). All seven patients diagnosed with gynecomastia at the beginning of the trial had grade 1 gynecomastia.
The patient assigned to bicalutamide and placebo had monolateral gynecomastia and, after 6 months, developed bilateral gynecomastia. Of the five patients affected by gynecomastia in the bicalutamide-tamoxifen group, three improved, and two remained stable. The patient assigned to bicalutamide-anastrozole remained stable.
PSA Response
Obviously, this analysis excluded the patients for whom only the baseline value was available (total, 12 patients; three, four, and five patients in the bicalutamide, bicalutamide-tamoxifen, and bicalutamide-anastrozole groups, respectively) and those patients with unmeasurable PSA levels (PSA < 0.2 ng/mL) at the beginning (total, seven patients; two, three, and two patients in the bicalutamide, bicalutamide-tamoxifen, and bicalutamide-anastrozole groups, respectively). Therefore, 94 patients (35, 30, and 29 patients in the bicalutamide, bicalutamide-tamoxifen, and bicalutamide-anastrozole groups, respectively) were included in the PSA analysis.
There was no major difference among groups in the proportion of patients achieving a ≥ 50% reduction in baseline PSA level, irrespective of the baseline cutoff value selected for this analysis (Table 2). However, a nonsignificant effect favored patients assigned to bicalutamide plus placebo or bicalutamide-tamoxifen compared with patients in the bicalutamide-anastrozole group when a ≥ 80% reduction in baseline PSA level was chosen for the analysis. This effect was maintained when results were adjusted by disease stage or pretreatment PSA level (data not shown).
Treatment Safety
Overall, treatments were well tolerated, with only 10%, 11%, and 14% of patients in the bicalutamide, bicalutamide-tamoxifen, and bicalutamide-anastrozole groups, respectively, experiencing serious adverse events. However, the incidence of adverse events was higher in the bicalutamide-anastrozole group when the analysis included all events, irrespective of severity (Table 3).
QOL and Sexual Functioning
This analysis was restricted to the patients who had answered the questionnaires administered after 6 months of treatment. Although there were no major differences among groups in mean sexual interest scores, there was just a borderline (P = .05) difference relative to sexual functioning scores. Indeed, scores increased in the bicalutamide plus placebo and bicalutamide-anastrozole groups, whereas they remained unchanged in the men assigned to bicalutamide-tamoxifen treatment (Table 4).
Pharmacodynamics
Again, this analysis was restricted to the patients for whom 6-month evaluations were available (Fig 4). Testosterone levels increased in all groups. However, there was no significant difference in the changes from baseline to month 6 among groups. The behavior of free testosterone in the patients assigned to bicalutamide and placebo or bicalutamide-anastrozole paralleled that of testosterone. In the patients assigned to tamoxifen, free testosterone levels remained substantially unchanged. The difference in change across time was statistically significant (P = .045). As expected, there was an increase in 17-estradiol levels in the bicalutamide and bicalutamide-tamoxifen groups, whereas there was a slight decrease in the levels of this hormone in the patients assigned to anastrozole. Overall, the difference in change across time was highly significant (P = .009).
Sex hormone–binding globulin levels sharply increased in the patients assigned to tamoxifen, whereas they decreased in patients assigned to placebo and remained substantially unchanged in the men assigned to anastrozole. Overall, the difference in change across time was highly significant (P = .006).
DISCUSSION
Tamoxifen was extremely effective in preventing the development of bicalutamide-induced gynecomastia and breast pain and was much more effective than anastrozole. These findings are in agreement with those reported by Saltzstein et al,13 whose prior study randomized 107 patients, who were receiving bicalutamide 150 mg as adjuvant to local treatment, to concurrently receive placebo, tamoxifen, or anastrozole for 3 months at the same daily doses used in the present study. At the end of 3 months of treatment, gynecomastia and breast pain had developed in 69% of patients assigned to placebo, 64% of patients assigned to anastrozole, and only 12% of patients assigned to tamoxifen, as assessed by physical examination and questionnaire.
In this study, patients on placebo and those initially assigned to tamoxifen or anastrozole who developed gynecomastia after discontinuation of medication were re-treated with tamoxifen or anastrozole for another 3 months. An objective improvement of gynecomastia was achieved in 65% and 80% of the patients initially assigned to placebo or tamoxifen and re-treated with this antiestrogen and 20% of patients re-treated with anastrozole13 (on file, AstraZeneca, 2002).
In our trial, five patients assigned to tamoxifen were found to have gynecomastia at the time of randomization. Three of these patients improved during treatment with tamoxifen. Despite small numbers, our trial findings support previous ones and confirm that tamoxifen is likely to be effective not only in the prophylaxis but also in the treatment of gynecomastia.
This trial was not specifically designed to compare the efficacy of the combinations on study with that of bicalutamide monotherapy. Indeed, this would require a much larger number of patients and a much longer follow-up time. Moreover, survival would represent a more appropriate end point. Nonetheless, it is reassuring that no major differences were recorded in PSA response rates, in particular between the patients assigned to bicalutamide and those assigned to bicalutamide-tamoxifen, irrespective of the cutoff value chosen to define response (≥ 50% v ≥ 80%) and of tumor stage and PSA baseline level.
Both tamoxifen and anastrozole have well-documented side effects that might be additive to those produced by bicalutamide 150 mg monotherapy, which would adversely affect treatment safety and tolerability. It is reassuring that few serious adverse events were reported in the present study and that they were independent of randomized treatment. Present data also suggest that bicalutamide 150 mg plus tamoxifen may have a better safety profile than bicalutamide 150 mg plus anastrozole and that the addition of tamoxifen to bicalutamide 150 mg does not increase the number and severity of adverse events.
It has been previously shown that bicalutamide 150 mg monotherapy can allow patients to preserve a better QOL. Four studies comparing bicalutamide 150 mg monotherapy with either castration or maximal androgen blockade and using the same 30-item questionnaire have shown that bicalutamide 150 mg monotherapy has benefits with respect to physical capacity, social functioning, emotional well-being, vitality, and sexual functioning.1-5
Although a detailed QOL analysis will form the object of a separate report, it can be anticipated that the addition of either tamoxifen or anastrozole to bicalutamide had no major detrimental effect on QOL (data not shown). In particular, the addition of tamoxifen did not worsen sexual interest and functioning. This is reassuring in view of a previous reporting of 29% sexual impotence in men treated with tamoxifen for breast cancer.11
The results of all pharmacodynamic assessments will be reported in full elsewhere. Here, the findings for those hormones known to be involved both in the development of gynecomastia and in the control of prostate cancer tumor growth are discussed. The present study confirms that antiandrogen monotherapy causes an increase in testosterone and free testosterone levels and that this surge is paralleled by a comparable increase in 17-estradiol levels because of the peripheral aromatization of testosterone.11 Our findings also confirm that tamoxifen and anastrozole are likely to affect the development of gynecomastia in different ways. The fact that 17-estradiol levels remained substantially unchanged in the men assigned to bicalutamide-anastrozole, whereas they increased in men assigned to bicalutamide alone, confirms that the activity of anastrozole on the breast is mainly related to its ability to interfere with 17-estradiol synthesis.
By contrast, a peripheral effect of tamoxifen at the receptor level is suggested by the high efficacy shown by this drug despite such a sharp increase in 17-estradiol levels. However, because a parallel increase of sex hormone–binding globulin occurred in this treatment group and because of the high binding affinity of this protein for circulating 17-estradiol, it cannot be ruled out that tamoxifen might work by reducing the bioavailability of estrogens as well. Unfortunately, we had not planned to dose free 17-estradiol and were not able to document an effect comparable to the effect achieved by bicalutamide-tamoxifen on free testosterone levels.
Whatever the mechanisms involved, other endocrine changes are likely to be involved in the development of gynecomastia beyond the progressive increase in 17-estradiol levels. It is noteworthy in this regard that the levels of prolactin, which is known to exert a permissive effect on estrogen activity at breast tissue level,18 were slightly increased by the administration of bicalutamide-anastrozole, whereas they were significantly lowered by the administration of bicalutamide-tamoxifen (data not shown). In view of the mechanisms involved in the antiproliferative activity of bicalutamide, maintenance of free testosterone levels in the same range of baseline values in the bicalutamide-tamoxifen group was also reassuring in respect to treatment antitumor efficacy.
In conclusion, the results of our study should be considered with caution because they are still preliminary, and overinterpretation of the findings should be avoided. However, it is clear that bicalutamide-induced gynecomastia and breast pain can be prevented by the concurrent administration of tamoxifen; anastrozole is much less effective in this regard. This beneficial effect can be achieved without altering PSA response rates and safety and without compromising sexual functioning. Pharmacodynamic studies provide reassurance on the putative hazard resulting from the early testosterone surge induced by bicalutamide 150 mg plus tamoxifen because no change in free testosterone was observed. However, safety and efficacy issues need to be investigated through more appropriate trials.
Authors' Disclosures of Potential Conflicts of Interest
Acknowledgment
We thank Simona Barozzi for her assistance in data management and secretarial support and Matteo Puntoni, BS, for cooperating in statistical analysis. We acknowledge the assistance of Vanasia (Milan, Italy) for data managing and Opera CRO (Genoa, Italy) for data monitoring. Finally, we thank AstraZeneca (Milan, Italy) for funding the trial and allowing us full autonomy and responsibility in designing the study and analyzing and interpreting the data.
NOTES
Presented in part at the 39th Annual Meeting of the American Society of Clinical Oncology, Chicago, IL, May 31-June 3, 2003.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Iversen P, Tyrrell CJ, Kaisary AV, et al: Casodex (bicalutamide) 150-mg monotherapy compared with castration in patients with previously untreated nonmetastatic prostate cancer: Results from two multicenter randomized trials at a median follow-up of 4 years. Urology 51:389-396, 1998
2. Iversen P, Tyrrell CJ, Kaisary AV, et al: Bicalutamide monotherapy compared with castration in patients with nonmetastatic locally advanced prostate cancer: 6.3 years of followup. J Urol 164:1579-1582, 2000
3. Tyrrell CJ, Kaisary AV, Iversen P, et al: A randomised comparison of ‘Casodex’ (bicalutamide) 150 mg monotherapy versus castration in the treatment of metastatic and locally advanced prostate cancer. Eur Urol 33:447-456, 1998
4. Boccardo F, Rubagotti A, Barichello M, et al: Bicalutamide monotherapy versus flutamide plus goserelin in prostate cancer patients: Results of an Italian Prostate Cancer Project study. J Clin Oncol 17:2027-2038, 1999
5. Chatelain C, Fourcade RO, Delchambre J: Bicalutamide (Casodex) versus combined androgen blockade (CAB): Open French multicentre study in patients with metastatic prostate cancer. Br J Urol 80:283, 1997 (abstr 1111)
6. See WA, McLeod D, Iversen P, et al: The bicalutamide Early Prostate Cancer Program: Demography. Urol Oncol 6:43-47, 2001
7. See WA, Wirth M, McLeod DG, et al: Bicalutamide as immediate therapy, either alone or as adjuvant to standard care of patients with localized or locally advanced prostate cancer: First analysis of the Early Prostate Cancer Program. J Urol 168:429-435, 2002
8. Tyrrell CJ: Tolerability and quality of life aspects with the anti-androgen Casodex (ICI 176,334) as monotherapy for prostate cancer: International Casodex Investigators. Eur Urol 26:15-19, 1994
9. McLeod DG, Iversen P: Gynecomastia in patients with prostate cancer: A review of treatment options. Urology 56:713-720, 2000
10. Decensi A, Torrisi R, Fontana V, et al: Long-term endocrine effects of administration of either a non-steroidal antiandrogen or a luteinizing hormone-releasing hormone agonist in men with prostate cancer. Acta Endocrinol (Copenh) 129:315-321, 1993
11. Anelli TF, Anelli A, Tran KN, et al: Tamoxifen administration is associated with a high rate of treatment-limiting symptoms in male breast cancer patients. Cancer 74:74-77, 1994
12. Boccardo F, Pace M, Rubagotti A, et al: Goserelin acetate with or without flutamide in the treatment of patients with locally advanced or metastatic prostate cancer: The Italian Prostatic Cancer Project (PONCAP) Study Group. Eur J Cancer 29A:1088-1093, 1993
13. Saltzstein D, Cantwell A, Sieber P, et al: Prophylactic tamoxifen significantly reduces the incidence of bicalutamide-induced gynecomastia and breast pain. Br J Urol 90:120-121, 2002 (abstr PD-4.07)
14. Armitage P, Berry G: Statistical Methods in Medical Research (ed 3). Oxford, United Kingdom, Blackwell, 1994
15. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
16. Peto R, Pike MC, Armitage P, et al: Design and analysis of randomized clinical trials requiring prolonged observation of each patient: II. Analysis and examples. Br J Cancer 35:1-39, 1977
17. Healy MJR: Some problems of repeated measurements, in Bithell JF, Coppi R (eds): Perspectives in Medical Statistics. London, United Kingdom, Academic Press, 1981, pp 155-171
18. Melis GB, Paoletti AM, Petacchi FD, et al: Physiological development, benign diseases and cancer of human breast: Importance of prolactin secretion, in Melis GB (ed): Postmenopausal Hormonal Therapy: Benefits and Risk. New York, NY, Raven Press, 1987, pp 55-65(F. Boccardo, A. Rubagotti)