Bone mineral density in Japanese prostate cancer patients under androgen-deprivation therapy

Transcription

1 Endocrine-Related Cancer (2008) Bone mineral density in Japanese prostate cancer patients under androgen-deprivation therapy Wei Wang*, Takeshi Yuasa, Norihiko Tsuchiya, Shinya Maita, Teruaki Kumazawa, Takamitsu Inoue, Mitsuru Saito, Zhiyong Ma, Takashi Obara, Hiroshi Tsuruta, Shigeru Satoh and Tomonori Habuchi Department of Urology, Akita University School of Medicine, Hondo, Akita , Japan (Correspondence should be addressed to T Yuasa; yuasa@doc.med.akita-u.ac.jp) *W Wang is now at Department of Urology, Second Affiliated Hospital of SunYat-sen University, Guangzhou, Guangdong province, China Abstract Androgen-deprivation therapy (ADT) of patients with prostate cancer (PCa) is known to reduce bone mineral density (BMD). However, the most studies examined Caucasian or black patients and the effects of ADT on the bone metabolism of East Asians are unclear. Therefore, we performed a crosssectional study to elucidate the influence of ADT on bone metabolism in Japanese patients. In total, 101 native Japanese patients with PCa were enrolled. They consisted of 58 ADT-treated and 43 hormone-naive patients. The BMD in the lumbar spine, total hip, and femoral neck was measured by dual energy X-ray absorptiometry and expressed in S.D. units relative to young adult men (T-score) or age-matched men (Z-score). Serum levels of bone metabolism markers were also measured. The BMDs at the three sites revealed that 2.3% (1/43) and 8.6% (5/58) of the hormone-naive and ADT-treated PCa patients had osteoporosis respectively, but this difference failed to achieve statistical significance (PZ0.294). The two groups also did not differ significantly in their Z-scores of the three sites, and univariate and multivariate analyses indicated that ADT was not a significant risk factor for decreased BMD. In addition, a significant correlation between the duration of ADT and BMD was not observed for all three sites measured. However, the ADT-treated patients had significantly higher serum levels of N-terminal telopeptide of type I collagen (NTx) than the hormone-naive patients (PZ0.017). To our knowledge, this is the first study to demonstrate the low prevalence of osteoporosis in both ADT-treated and hormone-naive Japanese PCa patients. Moreover, ADT did not significantly increase the prevalence of osteoporosis in this Japanese population. Endocrine-Related Cancer (2008) Introduction Androgen-deprivation therapy (ADT) has been the standard treatment for metastatic prostate cancer (PCa), since Huggins & Hodges first reported it in 1941 (Huggins & Hodges 1941). Currently, ADT is used not only to treat metastatic PCa, it is also frequently employed to treat non-metastatic PCa, where it is used either as a mainstream treatment for aging men with local disease or as a treatment for patients with biochemical failure after prostatectomy or radiation therapy (Cooperberg et al. 2003, Sharifi et al.2005, Nelson 2006). In addition, early ADT benefits patients with nodal metastases who have undergone prostatectomy and lymphadenectomy when compared with patients who received ADT later (Messing et al. 2006). Thus, patients are now frequently treated with ADT more often or for longer periods than in the past, which suggests that the adverse effects associated with ADT may become a greater clinical problem in the future. This necessitates more careful delineation and monitoring of the complications of ADT. ADT is associated with anemia, weight gain, insulin resistance, ischemic heart disease, hypogonadism, and increasing bone resorption (Higano 2003, Nelson 2006). Of these ADT complications, increasing bone resorption is of particular concern, because it may lead to osteoporosis and bone fractures (Smith 2004, Shahinian Endocrine-Related Cancer (2008) /08/ q 2008 Society for Endocrinology Printed in Great Britain DOI: /ERC Online version via

2 W Wang et al.: BMD in Japanese prostate cancer patients et al.2005). Indeed, a number of studies have shown that ADT-treated patients with PCa suffer from bone loss and skeletal-related adverse effects (Greenspan et al. 2005, Bruder et al.2006, Morote et al.2006, 2007), and a recent clinical report disclosed a negative correlation between skeletal fractures and overall survival in ADT-treated patients with PCa (Oefelein et al. 2002). These observations have led to the suggestion that clinicians should be alert to the impact of ADT on bone mineral density (BMD) and should strive to prevent bone loss. One limitation of the studies examining the relationship between ADT treatment and BMD is that most focused on Caucasian or black patients with PCa. Several studies have shown that Japanese and Caucasian women show racial differences in BMD and the incidence of bone fracture (Shimizu et al. 1990, Ito et al. 1997, Yoneda et al. 2006). One of these studies revealed that Japanese women have a lower BMD and a lower threshold of bone fracture than Caucasian women (Ito et al. 1997). These racial differences between Caucasian and Japanese women were also highlighted by a breast cancer study that showed that aromatase inhibitor treatment of Japanese women only slightly reduced their estrogen levels whereas Caucasian women showed a large reduction in estrogen levels; thus, aromatase inhibitor treatment is less likely to affect bone turnover and BMD levels in Japanese women than in Caucasian women (Yoneda et al. 2006). These observations suggest that Japanese men could differ from Caucasian men in terms of the influence of ADT on BMD and bone metabolism. To test this notion, we here examined the effect of ADT on bone metabolism in Japanese patients with PCa. Materials and methods Subjects A total of 101 native Japanese patients with PCa were enrolled in this study. These patients comprised 58 ADT-treated and 43 hormone-naive patients who were treated at the Akita University Medical Center from December 2006 to November Of the 58 ADT-treated patients, 37 and 21 did and did not have bone metastasis respectively. Patients with bone metastases in the lumbar spine or the hip joint were not included in this study as these were the regions we subjected to BMD measurement. The subjects completed a baseline questionnaire regarding life-style factors, including caffeinated beverage servings per day, smoking status, packs per year of cigarettes smoked, alcoholic drinks per week, use of calcium and vitamin D supplements, and exercise (days per week and total hours per week). All PCa patients were diagnosed on the basis of histological analysis of specimens obtained from transrectal needle biopsy or transurethral resection of the prostate for voiding symptoms. All 58 patients who underwent surgical castration or received a luteinizing hormone-releasing hormone agonist were demonstrated to have castrated levels of serum testosterone. In addition, 48 of the 58 ADT-treated patients (82.8%) also received bicalutamide (80 mg/day). The pathological grading of PCa was determined according to Gleason s histological grading and the tumor node metastatic system (Gleason 1977, Sobin & Wittekind 2002). BMD measurements BMD was measured in our hospital in by dual energy X-ray absorptiometry (DXA) using a Delphi QDR (Hologic, Bedford, MA, USA). The area of BMD in grams per square centimeter was measured at the posteroanterior spine (L2 L4) and the non-dominant hip (the total hip and the femoral neck). Peak BMD, age-specific BMD, peak S.D., and age-specific S.D. for DXA values were derived from the Hologic database for East Asian ethnicity (version 2.0; Hologic, Inc). The coefficient of variation was less than 1%. BMD was expressed in S.D. units relative to young adult men (T-score) and relative to age-matched men (Z-score). According to World Health Organization criteria, a normal BMD is defined as a T-score greater than K1 S.D.s, osteopenia as a T-score between K1 and K2.5, and osteoporosis as a T-score of K2.5 or less (Miller 2006). Measurement of biochemical values The serum levels of N-terminal telopeptide of type I collagen (NTx, normal range nmol/l), C-terminal telopeptides of type I collagen (ICTP, normal range ng/ml), intact parathyroid hormone (intact PTH, normal range pg/ml), prostate-specific antigen (PSA, normal range less than 4 ng/ml), and testosterone (normal range ng/ml) were measured before breakfast at the time BMD was measured. The intraassay coefficients of variation for NTx and ICTP were less than 10 and 8% respectively. In addition, common laboratory blood and serum data, including hemoglobin (Hb), albumin (Alb), lactate dehydrogenase (LDH), alkaline phosphatase (ALP), Ca, P, and glucose (Glu) levels, were determined using a multichannel autoanalyzer (LX-20, Beckman-Coulter, Los Angeles, CA, USA). Statistical analysis Differences in the clinical and serum variables of the hormone-naive patients, the ADT-treated patients 944

3 Endocrine-Related Cancer (2008) without bone metastasis, and the ADT-treated patients with bone metastasis were evaluated using the Student s t-test. Alternatively, the Mann Whitney U test was used if the group variances were unequal or the dependent variables were non-normally distributed and not transformable. The groups were also compared with regard to the prevalences of normal BMDs, osteopenia, osteoporosis, and life-style factors and analyzed statistically using the Kruskal Wallis test. Univariate analysis was used to assess the association between BMD and age, body mass index (BMI), serum PSA levels, presence or absence of ADT, and duration of ADT. Multivariate backward stepwise linear regression analysis was performed to identify the parameters that were independently and significantly associated with BMD. Nonparametric Spearman rank-order correlation coefficients were calculated to investigate the relationship between the BMD, T-score, and Z-score and a variety of bone metabolic markers. A modest positive correlation was considered to be an r value of , whereas a strong correlation was considered to be an r value greater than 0.5. All statistical analyses were performed using the Statistical Package for Social Sciences (SPSS version 13.0; SPSS Inc, Chicago, IL, USA) and two-sided P values less than 0.05 were considered to indicate statistical significance. Results Patient characteristics The 101 patients were divided into hormone-naive patients (nz43), ADT-treated patients without bone metastasis (nz37), and ADT-treated patients with bone metastasis (nz21) and their characteristics, including their age, BMI, biopsy Gleason score, serum or blood levels of Alb, Hb, PSA and testosterone, and the presence and duration of ADT, are summarized in Table 1. In addition, the life-style factors of these patients, including caffeinated beverage consumption, smoking status, alcoholic drink consumption, use of calcium and vitamin D supplements, and exercise, are summarized in Table 1. The ADT-treated patients without bone metastasis were significantly older than the other two groups, while the the ADT-treated patients with bone metastasis tended to have higher serum PSA values than those without bone metastasis (PZ0.053). The ADT-treated patients without bone metastasis tended to have higher serum LDH values than the hormone-naive patients (PZ0.054), while the hormone-naive patients had significantly higher serum levels of testosterone than the two ADT-treated groups. However, all three groups were similar in terms of the other variables that were recorded, including BMI, Gleason score, Hb and serum levels of Alb and Glu (Table 1). They also did not differ significantly in terms of the life-style factors that were recorded, including caffeinated beverage consumption, smoking status, packs per year of cigarettes, alcohol consumption, calcium or vitamin D supplementation, and exercise (Table 1). Comparison of the BMDs, T-scores, and Z-scores of the hormone-naive and ADT-treated patients According to new guidelines, osteoporosis can be diagnosed when DXA of the lumbar spine, total hip, or femoral neck yields a T-score of K2.5 or less (Miller 2006). Therefore, we measured the BMD at these three sites in this study. Since bone metastatic lesions can affect the BMD results, we initially excluded the bone metastatic patients and compared the BMDs, T-scores, and Z-scores of the hormone-naive patients and the ADT-treated patients without bone metastasis (Table 2). In addition, when calculating T-score, the International Society of Clinical Densitometry recommends to use a uniform Caucasian (non-race adjusted) male normative database for men of all ethnic groups, although they also described that the application of recommendation may vary according to the local requirement. On current, there is no consensus of the database in calculating T-score of the patients under ADT. Therefore, we used both databases when calculating T-scores in this study. The BMD values and T-scores at the three sites of the ADT-treated patients were all lower than those of the hormone-naive patients, although these differences did not achieve statistical significance (Table 2). In the International Society of Clinical Densitometry recommendation, the patient s self-reported ethnicity should be used when calculating Z-scores. The average Z-scores (which indicate the S.D. units compared with age-matched normal relatives) of the ADT-treated patients and the hormone-naive patients were positive in this study. Prevalence of osteoporosis and osteopenia in the hormone-naive and ADT-treated patients We compared the prevalence of osteoporosis and osteopenia in the hormone-naive patients and the ADT-treated patients without bone metastasis (Table 3). At first, in order to decide the prevalence of osteopenia and osteoporosis, we calculated T-scores using East Asian database (Table 3). When focusing on the worst site showing low T-score, the prevalence of osteoporosis and osteopenia for the hormone-naive PCa patients was 2.3% (1 patient) and 44.2% (19 patients) respectively. By contrast, for the ADT-treated PCa patients, the respective prevalence was 10.8% (4 patients) and 40.6% (15 patients). These differences were not 945

4 946 Table 1 Comparison of the characteristics and lifestyle factors of the three groups of patients with prostate cancer Under androgen-deprivation therapy (ADT) Hormone-naive Bone metastasis (K) (nz43) Bone metastasis (K) (nz37) Bone metastasis (C) (nz21) P a P b (A) Comparison of the characteristics of the three groups of patients with prostate cancer Age 68.1G G G Body mass index (kg/m 2 ) 23.4G G G Biopsy Gleason score 7.2G G G PSA (ng/ml) 10.3G G G Hemoglobin (g/dl) 12.6G G G Albumin (g/dl) 4.1G G G Glucose (g/dl) 112.8G G G LDH (IU/l) 171.7G G G Testosterone (ng/ml) 4.1G G G !10 K Duration of ADT (month) 23.5G G (B) Comparison of the life-style factors of the three groups of patients with prostate cancer Caffeinated beverages/day (40%) 14 (38%) 11 (52%) (49%) 16 (43%) 3 (14%) R3 5 (11%) 3 (8%) 2 (10%) Unknown 0 (0%) 4 (11%) 5 (24%) Smoking status Never 27 (63%) 18 (49%) 10 (47%) Former 12 (28%) 18 (49%) 6 (29%) Current 4 (9%) 1 (2%) 0 (0%) Unknown 0 (0%) 0 (0%) 5 (24%) Packs per year of cigarettes (63%) 18 (49%) 10 (47%)!20 14 (33%) 13 (35%) 6 (29%) R20 2 (4%) 6 (16%) 0 (0%) Unknown 0 (0%) 0 (0%) 5 (24%) Alcohol (drinks/week) (44%) 14 (38%) 8 (38%) (21%) 8 (22%) 2 (10%) R7 15 (35%) 15 (41%) 6 (29%) Unknown 0 (0%) 0 (0%) 5 (24%) Calcium supplements No 41 (95%) 31 (84%) 15 (71%) Yes 2 (5%) 6 (16%) 1 (5%) Unknown 0 (0%) 0 (0%) 5 (24%) Vitamin D supplements No 40 (93%) 33 (89%) 15 (71%) Yes 3 (7%) 4 (11%) 1 (5%) Unknown 0 (0%) 0 (0%) 5 (24%) W Wang et al.: BMD in Japanese prostate cancer patients

5 Endocrine-Related Cancer (2008) Table 1 continued Under androgen-deprivation therapy (ADT) Hormone-naive Bone metastasis (K) (nz43) Bone metastasis (K) (nz37) Bone metastasis (C) (nz21) P a P b Exercise (days/week) (37%) 18 (49%) 7 (33%) (30%) 8 (22%) 4 (19%) (33%) 9 (24%) 5 (24%) Unknown 0 (0%) 2 (5%) 5 (24%) Exercise (h/week) (37%) 18 (49%) 7 (33%)!2 8 (19%) 7 (19%) 4 (19%) R2 19 (44%) 10 (27%) 5 (24%) Unknown 0 (0%) 2 (5%) 5 (24%) Values indicate meangs.d. The data are presented as number of patients (percentages). a Probability values of the differences between the hormone-naive patients and the ADT-treated patients without bone metastasis. b Probability values of the differences between the ADT-treated patients with and without bone metastasis. significant. In addition, we also decided the prevalence of osteopenia and osteoporosis using Caucasian database. We found that the differences of the prevalence of osteopenia and osteoporosis were also not significant (Supplementary Table 1, which can be viewed online at These results suggest that the PCa patients without bone metastasis did not show a significant increase in the prevalence of osteoporosis and osteopenia after a relatively short period of ADT (average 23.5 months). Association of the ADT duration with BMD, T-score, and Z-score of the patients with non-metastatic PCa A study examining the BMD of Caucasian patients with PCa revealed that the greatest ADT-induced bone loss occurred in the first 12 months, after which time it showed a slower decrease (Bunker et al. 2006). To determine whether Japanese PCa patients showed a similar pattern, we investigated the association between the duration of ADT and BMD, the T-score, and the Z-score in ADT-treated patients with non-metastatic PCa (Fig. 1). For all three sites measured (L2 L4 spine, femoral neck, and total hip joint), a significant correlation between the duration of ADT and BMD was not observed (rzk0.152, PZ0.147; rzk0.020, PZ0.850; and rzk0.045, PZ0.670, respectively). To determine whether other patient characteristics were associated with the lower BMD in ADT-treated men, we performed a series of regression analyses using the data from the ADT-treated PCa patients without metastasis. BMD was weakly positively correlated with BMI (rz0.379, PZ0.001) and weakly inversely correlated with age (rzk0.337, PZ0.004). However, no other factors correlated significantly with the BMD at any of the three sites, including the serum PSA levels, the biopsy Gleason score, the presence or absence of ADT, and the duration of ADT. Furthermore, none of these factors, including ADT, were independent predictors in a multivariate linear regression model using the BMDs at the three sites. These observations together suggest that ADT does not have a big impact on BMD in Japanese patients. Influence of ADT on serum variables and bone metabolism To determine the effect of ADT on serum variables associated with bone metabolism, we compared the hormone-naive patients and the ADT-treated patients without metastasis with regard to these variables. As shown in Table 4, the ADT-treated patients had significantly higher serum levels of NTx, which reflects bone resorption, than the hormone-naive patients 947

6 W Wang et al.: BMD in Japanese prostate cancer patients Table 2 Comparison of the bone mineral density (BMD), T-score, and Z-score of the three groups of patients with prostate cancer Under androgen-deprivation therapy (ADT) Hormone naive Bone metastasis (K) Bone metastasis (K) Bone metastasis (C) P a P b Lumbar spine (PA) BMD 1.050G G G T-score (East Asian) c 0.037G1.603 K0.465G1.482 K0.300G T-score (Caucasian) d K0.573G2.082 K1.350G1.599 K1.300G Z-score 0.593G G G Femoral neck BMD 0.763G G G T-score (East Asian) c K0.793G0.900 K0.981G0.863 K0.829G T-score (Caucasian) d K1.206G0.669 K1.443G0.846 K1.313G Z-score 0.408G G G Total hip BMD 0.891G G G T-score (East Asian) c K0.512G1.066 K0.820G0.990 K0.619G T-score (Caucasian) d K0.918G0.792 K1.312G0.870 K1.035G Z-score 0.519G G G Values indicate meangs.d. a Probability values of the differences between the hormone-naive patients and the ADT-treated patients without bone metastasis. b Probability values of the differences between the ADT-treated patients with and without bone metastasis. c T-score calculated with East Asian database. d T-score calculated with Caucasian database. (PZ0.017) but the two groups did not differ significantly in any of the other serum variables. Next, we investigated the association between the bone mineral variables (BMD, T-score, Z-score) and the serum variables associated with bone metabolism (PSA, Ca, P, ALP, NTx, ICTP, and intact PTH). The serum NTx level was the only variable that was significantly associated with the BMDs of all three sites that were measured (L2 L4 spine, femoral neck, and total hip joint), although the correlations were relatively weak (rzk0.330, PZ0.005; rzk0.338, PZ0.004; and rzk0.434, PZ0.0001, respectively). Thus, while ADT is only weakly associated with BMD loss, it is significantly associated with increased serum NTx levels. Table 3 Comparison of the frequencies of osteoporosis and osteopenia of the three groups of patients with prostate cancer a Under androgen-deprivation therapy (ADT) Hormone-naive Bone metastasis Bone metastasis (K) Bone metastasis (C) P b P c Lumbar spine (PA) Normal 31 (72.1%) 22 (59.5%) 15 (71.4%) Osteopenia 11 (25.6%) 11 (29.7%) 5 (23.8%) Osteoporosis 1 (2.3%) 4 (10.8%) 1 (4.8%) Femoral neck Normal 23 (53.5%) 18 (48.6%) 13 (61.9%) Osteopenia 19 (44.2%) 18 (48.6%) 7 (33.3%) Osteoporosis 1 (2.3%) 1 (2.7%) 1 (4.8%) Total hip Normal 34 (79.1%) 22 (59.5%) 14 (66.7%) Osteopenia 8 (18.6%) 12 (32.4%) 6 (28.6%) Osteoporosis 1 (2.3%) 3 (8.1%) 1 (4.8%) Worst site Normal 23 (53.5%) 18 (48.6%) 13 (61.9%) Osteopenia 19 (44.2%) 15 (40.6%) 7 (33.3%) Osteoporosis 1 (2.3%) 4 (10.8%) 1 (4.8%) Values indicate the numbers (percentile). a T-scores were calculated with East Asian database. b Probability values of the differences between the hormone-naive patients and the ADT-treated patients without bone metastasis. c Probability values of the differences between the ADT-treated patients with and without bone metastasis

7 Endocrine-Related Cancer (2008) Figure 1 Association between the duration of androgen-deprivation therapy and the BMD, T-score, and Z-score of the lumbar spine, femoral neck, and total hip joints. This suggests that measuring NTx levels may be a better way of monitoring bone loss in Japanese patients. Comparison of the BMDs, T-scores, and Z-scores of the ADT-treated patients with and without bone metastasis Osteoblastic metastatic lesions, which are typical in patients with bone metastasis and PCa, may artificially increase the BMD score. To determine the extent to which the presence of bone metastasis influences the BMD, we compared the clinical parameters of the ADT-treated patients with and without bone metastasis (Tables 2 and 3). These two groups of patients did not differ significantly in BMDs, T-scores, or Z-scores, which suggest that the DXA-measured BMD values of patients with PCa are not disturbed by the presence of bone metastases. Moreover, the two groups of ADT-treated patients did not differ significantly in their serum NTx levels (Table 4), which indicates that the NTx serum levels reflect the effect of ADT on bone resorption and are not altered by the presence of bone metastases. Table 4 Comparison of the serum variables of the three groups of patients with prostate cancer Under androgen-deprivation therapy (ADT) Hormone naive Bone metastasis (K) Bone metastasis (K) Bone metastasis (C) P a P b ALP (IU/l) 227.5G G G Calcium (mg/dl) 8.9G G G Phosphate (mg/dl) 3.2G G G Intact PTH (pg/ml) 34.5G G G NTx (nmol/l) 13.3G G G ICTP (ng/ml) 5.1G G G Values indicate meangs.d. a Probability values of the differences between the hormone-naive patients and the ADT-treated patients without bone metastasis. b Probability values of the differences between the ADT-treated patients with and without bone metastasis

8 W Wang et al.: BMD in Japanese prostate cancer patients Discussion The effect of ADT on BMD and its ability to induce osteoporosis in Caucasians and blacks, mainly in western countries, are now well documented (Greenspan et al. 2005, Bruder et al. 2006, Morote et al. 2006, 2007). However, it remains unclear whether ADT has a similar influence on the BMD of East Asian patients with PCa. Although other studies found a high incidence of osteoporosis in western ADT-treated patients with PCa (Bruder et al. 2006, Morote et al. 2006, 2007), we found that this was not true for Japanese patients with PCa, as only 10.8% (4/37) of the ADT-treated patients exhibited osteoporosis. However, ADT did mildly promote osteoporosis in Japanese PCa patients, as even fewer of the hormone-naive PCa patients (2.3%; 1/43) had osteoporosis, although this difference from ADT-treated patients did not quite achieve statistical significance (PZ0.294). In addition, since there was no correlation between the duration of ADT and BMD or other parameters, our study suggested that longer durations of ADT do not significantly decrease the BMD of Japanese PCa patients and therefore may not increase the prevalence of osteoporosis in these patients. It is possible that we did not detect a significant difference in BMD between ADT-treated and hormone-naive PCa patients (PZ0.294) because the patients in this study only received ADT for a relatively short period (on average 23.5 months). However, many studies of Caucasian patients with PCa have shown high incidences of osteoporosis in patients receiving ADT for similarly short durations; moreover, even ADT-naive Caucasian patients show a high incidence of osteoporosis (Bruder et al. 2006, Morote et al. 2006, 2007). For example, Bruder et al. (2006) showed that of 89 patients with a median duration of ADT of 2.7G2.5 years, whose BMDs in the spine and total hip had been measured, 26.9 and 50.6% had osteoporosis and osteopenia respectively. Moreover, Morote et al. (2007) reported that 35.4 and 45.2% of hormone-naive patients had osteoporosis and osteopenia respectively (Morote et al. 2007), while 42.9 and 39.3% of patients that had been treated with ADT for 2 years suffered osteoporosis and osteopenia respectively. They also showed that the greatest bone loss occurred in the first year after the initiation of androgen deprivation (Morote et al. 2006). Thus, it is unlikely that the relatively short duration of ADT in our study explains why we did not detect a significant difference between hormone-naive and ADT-treated Japanese patients with PCa. Moreover, it appears that Caucasians have a higher baseline incidence of BMD loss and a higher susceptibility to ADT-induced bone loss than Japanese patients. Only two other studies have sought to evaluate the BMD in Japanese patients with PCa (Egawa et al. 2000, Miyaji et al. 2004). Miyaji et al. (2004) showed that 27 patients with PCa exhibited mild but significant bone loss after being treated for 2 years with ADT using gonadotropin-releasing hormone agonists, as their median BMD dropped from to g/cm 2 (PZ0.047). Egawa et al. (2000) reported a pilot study of 17 patients with PCa treated intermittently with ADT and stated that 10 (58.8%) patients maintained BMD levels in the normal range for their age. However, neither of these studies measured the rates of osteoporosis in these patients. Thus, we show here for the first time that Japanese PCa patients suffer low rates of osteoporosis regardless of whether they are treated with ADT. We also observed that the average Z-score of the patients under ADT was not lower than that of hormone-naive patients, and that these average Z-scores of the hormone-naive patients and the ADT-treated patients with and without bone metastasis were positive. These results indicate that the average BMDs of the patients in this study were higher than that of age-matched East Asian controls. While this could reflect a selection bias, it could also suggest that Japanese men with high baseline BMDs have a higher propensity to develop PCa. This notion is supported by a recent report showing that high bone density in older Afro-Caribbean men is associated with PCa (Bunker et al. 2006). However, other studies in western countries have not observed this trend (Bruder et al. 2006, Morote et al. 2006, 2007). The reasons for these differences between Caucasians, westerners, and Japanese are not clear but may reflect racial differences in the genetic factors that determine the BMD. Treatment and life-style factors may also be important determinants that drive these differences. Another important finding of this study was that the mean serum NTx level of non-metastatic patients under ADT was significantly higher than that of hormone-naive patients (PZ0.017). Serum NTx levels reflect bone resorption and our observation is consistent with our other finding that ADT treatment of Japanese PCa patients mildly increases the loss of BMD, although this difference did not achieve statistical significance (PZ0.294). Notably, we also showed that the NTx serum levels of the ADT-treated patients correlated weakly but significantly with BMD (rz0.379, PZ0.001). These observations suggest that serum NTx can reflect the impact of ADT on bone resorption more precisely than BMD and that detection 950

9 Endocrine-Related Cancer (2008) of an increase in serum NTx in ADT patients may herald a significant decrease in bone mass. Further investigations are needed to confirm the correlation of NTx and BMD under ADT. It should be noted that ADT and bone metastasis frequently coexist in PCa patients and that osteoblastic metastatic lesions may artificially increase the BMD. Supporting this is that regions with osteoblastic bone metastasis show significantly higher bone density than regions with osteolytic bone lesions (Vassiliou et al. 2007). These observations have led many BMD studies to exclude patients with bone metastasis. It also led us to exclude patients with hot spots in the lumbar spine or the hip joint (which are the bone regions we evaluated in this study) and to separate the ADT-treated patients with bone metastasis from those without bone metastasis. Significantly, however, we did not detect any differences in BMD or serum NTx levels between ADT-treated PCa patients with or without bone metastasis. This is consistent with a report by Michaelson et al. (2004) that showed the serum NTx levels of ADT-treated PCa patients were significantly higher than those of hormone-naive men (P!0.01) but did not differ when bone metastasis was present (PZ0.33). Thus, the effects of bone metastases on bone metabolism do not seem to disturb the DXA measurement of BMD or the serum NTx values. The presence of bone metastases also did not perturb other serum variables reflecting bone metabolism. That the patients with and without bone metastasis had comparable BMDs is important. One of the main reasons for measuring the BMD of patients with PCa is to enhance the early detection of pathological BMD loss, which allows appropriate treatment for osteoporosis to be instituted, which in turn reduces the incidence of bone fractures. Patients with bone metastases have the highest risk for bone fracture because they bear many risk factors, including older age, treatment with ADT, corticosteroid use, and bone metastasis (Higano 2003, Nelson 2006). Our observations suggest that we may be able to use DXA to evaluate the BMD of PCa patients with bone metastasis if it is properly conducted. In conclusion, our study showed that Japanese PCa patients did not have high rates of low BMD, regardless of whether they were hormone-naive or treated with ADT. We also did not detect an association between the length of ADT and BMD, T-score, or Z-score. These findings are quite different from previous studies examining patients in western countries. To our knowledge, this is the first study to demonstrate the low rates of osteoporosis in both ADT-treated and hormone-naive Japanese PCa patients. Given the cross-sectional nature of this study and its relatively small sample size, it will be necessary to conduct further larger scale and prospective studies to confirm our observations and to determine the genetic background and life-style factors that influence the BMD of PCa patients. Declaration of interest The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported. Funding This work was partly supported by Kobayashi Institute for Innovative Cancer Chemotherapy, the Shimadzu Science Foundation, the Sagawa Foundation for Promotion of Cancer Research, the Japan-China Sasakawa Medical Fellowship, Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan, and the GCOE program of the Ministry of Education, Culture, Sports, Science and Technology, Japan. References Bruder JM, Ma JZ, Basler JW & Welch MD 2006 Prevalence of osteopenia and osteoporosis by central and peripheral bone mineral density in men with prostate cancer during androgen-deprivation therapy. Urology Bunker CH, Zmuda JM, Patrick AL, Wheeler VW, Weissfeld JL, Kuller LH & Cauley JA 2006 High bone density is associated with prostate cancer in older Afro-Caribbean men: Tobago prostate survey. Cancer Causes and Control Cooperberg MR, Grossfeld GD, Lubeck DP & Carroll PR 2003 National practice patterns and time trends in androgen ablation for localized prostate cancer. Journal of the National Cancer Institute Egawa S, Takashima R, Matsumoto K, Mizoguchi H, Kuwao S & Baba S 2000 A pilot study of intermittent androgen ablation in advanced prostate cancer in Japanese men. Japanese Journal of Clinical Oncology Gleason DF 1977 Histological grading and clinical staging of prostate carcinoma. In Urologic Pathology, pp Ed M Tannenbaum. Philadelphia: Lea & Feiger. Greenspan SL, Coates P, Sereika SM, Nelson JB, Trump DL & Resnick NM 2005 Bone loss after initiation of androgen deprivation therapy in patients with prostate cancer. Journal of Clinical Endocrinology and Metabolism Higano CS 2003 Side effects of androgen deprivation therapy: monitoring and minimizing toxicity. Urology

10 W Wang et al.: BMD in Japanese prostate cancer patients Huggins C & Hodges VC 1941 Studies on prostate cancer: I. The effect of castration, of estrogen and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate. Cancer Research International Union Against Cancer 2002 Urological tumors: prostate. In TNM Classification of Malignant Tumors, edn 6, pp Eds LH Sobin & C Wittekind. New York: Wiley-Liss. Ito M, Lang TF, Jergas M, Ohki M, Takada M, Nakamura T, Hayashi K & Genant HK 1997 Spinal trabecular bone loss and fracture in American and Japanese women. Calcified Tissue International Messing EM, Manola J, Yao J, Kiernan M, Crawford D, Wilding G, di SantAgnese PA, Trump D & Eastern Cooperative Oncology Group study EST Immediate versus deferred androgen deprivation treatment in patients with node-positive prostate cancer after radical prostatectomy and pelvic lymphadenectomy. Lancet Oncology Michaelson MD, Marujo RM & Smith MR 2004 Contribution of androgen deprivation therapy to elevated osteoclast activity in men with metastatic prostate cancer. Clinical Cancer Research Miller PD 2006 Guidelines for the diagnosis of osteoporosis: T-scores vs fractures. Reviews in Endocrine and Metabolic Disorders Miyaji Y, Saika T, Yamamoto Y, Kusaka N, Arata R, Ebara S, Nasu Y, Tsushima T & Kumon H 2004 Effects of gonadotropin-releasing hormone agonists on bone metabolism markers and bone mineral density in patients with prostate cancer. Urology Morote J, Orsola A, Abascal JM, Planas J, Trilla E, Raventos CX, Cecchini L, Encabo G & Reventos J 2006 Bone mineral density changes in patients with prostate cancer during the first 2 years of androgen suppression. Journal of Urology Morote J, Morin JP, Orsola A, Abascal JM, Salvador C, Trilla E, Raventos CX, Cecchini L, Encabo G & Reventos J 2007 Prevalence of osteoporosis during long-term androgen deprivation therapy in patients with prostate cancer. Urology Nelson JB 2006 Hormonal therapy for prostate cancer. In Campbell-Walsh UROLOGY, edn 9, pp Eds AJ Wein, LR Kavoussi, AC Novick, AW Partin & CA Peters. New York: Elsevier. Oefelein MG, Ricchiuti VS, Conrad PW, Goldman H, Bodner D, Resnick MI & Seftel A 2002 Clinical predictors of androgen-independent prostate cancer and survival in the prostate-specific antigen era. Urology Shahinian VB, Kuo YF, Freeman JL & Goodwin JS 2005 Risk of fracture after androgen deprivation for prostate cancer. New England Journal of Medicine Sharifi N, Gulley JL & Dahut WL 2005 Androgen deprivation therapy for prostate cancer. Journal of the American Medical Association Shimizu H, Ross RK, Bernstein L, Pike MC & Henderson BE 1990 Serum oestrogen levels in postmenopausal women: comparison of American whites and Japanese in Japan. British Journal of Cancer Smith MR 2004 Osteoporosis and obesity in men receiving hormone therapy for prostate cancer. Journal of Urology 172 S52 S56. Vassiliou V, Kalogeropoulou C, Petsas T, Leotsinidis M & Kardamakis D 2007 Clinical and radiological evaluation of patients with lytic, mixed and sclerotic bone metastases from solid tumors: is there a correlation between clinical status of patients and type of bone metastases? Clinical & Experimental Metastasis Yoneda K, Tanji Y, Okishiro M, Taguchi T, Tamaki Y & Noguchi S 2006 Influence of adjuvant anastrozole on bone mineral density in Japanese postmenopausal breast cancer patients: is there a racial difference? Annals of Oncology