The use of selective estrogen receptor modulators on bone health in men

Abstract

Selective estrogen receptor modulators (SERMs) represent a class of drugs that act as agonist or antagonist for estrogen receptor in a tissue-specific manner. The SERMs drugs are initially used for the prevention and treatment of osteoporosis in postmenopausal women. Bone health in prostate cancer patients has become a significant concern, whereby patients undergo androgen deprivation therapy is often associated with deleterious effects on bone. Previous preclinical and epidemiological findings showed that estrogens play a dominant role in improving bone health as compared to testosterone in men. Therefore, this evidence-based review aims to assess the available evidence derived from animal and human studies on the effects of SERMs on the male skeletal system. The effects of SERMs on bone mineral density (BMD)/content (BMC), bone histomorphometry, bone turnover, bone strength and fracture risk have been summarized in this review.

Keywords:

• Osteoporosis
• tamoxifen
• raloxifene
• lasofoxifene
• toremifene

Introduction

Male osteoporosis is a serious but under-recognized public health problem because its prevalence is lower compared to female osteoporosis. Of the nine million fractures occurred in population aged 50 years and above in 2000, 39% happened in men. Specifically, 30% of hip, 20% of forearm, 42% of clinical vertebral and 25% of humerus fractures happened in men [1]. Additionally, similar trends were obtained from study by Liu et al. [2] demonstrated that the increase in age (20–59 years old) was associated with the decrease in cancellous bone microstructure, density and biomechanics in men. Studies also showed that mortality and risk of subsequent fractures were higher in elderly men compared to their women counterpart [3,4]. For instance, the Dubbo Osteoporosis Epidemiology Study revealed that the postfracture mortality rate was 2.05 [95% CI: 1.43–2.83] in women and 2.43 [95% CI: 1.95–2.99] in men aged 60 years and above [5].

Hypogonadism or testosterone deficiency in adult men can be defined as a clinical and biochemical syndrome associated with low testosterone, which may affect multiple organ functions and quality of life [6]. Hypogonadism is a major cause of male osteoporosis. It can be divided into primary and secondary hypogonadism. Primary hypogonadism occurs due to the incapacity of the testes to synthesize sufficient testosterone and is related to aging in men. Androgen production declines over lifetime, and approximately, 20% men older than 60 years have a serum testosterone level in the hypogonadal range [7,8]. Secondary hypogonadism is caused by the disturbance of hypothalamus–pituitary axis resulted from disease (pituitary cancer for example) or drugs [gonadotrophin-releasing hormone (GnRH) agonist used in prostate cancer for example] [9]. Aging men and patients undergoing androgen deprivation therapy using GnRH agonists are known to suffer from a higher rate of bone loss and increased fracture risk [10–12]. As testosterone is converted to estradiol by aromatase, lower total and bioavailable testosterone levels consequently translate to lower total and bioavailable estradiol levels [13].

The relative contribution of androgen and estrogen in the regulation of bone metabolism remains debatable to date. Testosterone acts on bone metabolism directly via binding to the androgen receptor (AR) or indirectly via aromatization to estradiol to activate estrogen receptor-α (ER-α) and/or estrogen receptor-β (ER-β) [14]. An in vivo study by Movérare et al. [15] investigated the differential effects of AR and ER-α activation on bone in orchidectomized (ORX) adult male mice. Results indicated that activation of ER-α preserved the trabecular number (Tb.N), trabecular thickness (Tb.Th), cortical thickness, and cortical density. In contrast, activation of AR only preserved Tb.N. These observations suggested that the bone-sparing effect of ER-α activation was clearly distinct from the effects of AR activation in adult gonadectomized male mice [15]. In human observational studies, estrogen was found to have an independent and stronger correlation with bone mineral density compared to testosterone [16,17].

Selective estrogen receptor modulators (SERMs) are a class of drugs that exert selective agonist or antagonist effects on various estrogen target tissues [18]. They have drawn considerable attention because they exert selective estrogen effects on target organs with minimal harmful side effects associated with hormone replacement therapy [19]. Tamoxifen and raloxifene are two of the best-characterized SERMs that act predominantly as estrogen antagonists in breast cancer cells, but agonists on bone cells [20]. Tamoxifen has been used for many years as an adjuvant therapy for hormone-responsive breast cancer [21], while raloxifene has been approved for the prevention and treatment of osteoporosis in postmenopausal women [22].

The deleterious effects of androgen deprivation therapy on bone health have become a significant concern for prostate cancer patients undergoing therapy. Evidence from the human studies showed that SERMs are effective in reducing skeletal adverse effects associated with prostate cancer treatment [23,24]. This evidence-based review summarized the effects of SERMs on bone health derived from animal and human studies in terms of bone mineral density (BMD)/content (BMC), bone histomorphometry, bone turnover, bone strength and fracture risk.

Literature search

A literature search was performed using PubMed and Scopus using keywords: “selective estrogen receptor modulators” AND “bone OR fracture” AND “men OR male”. The search was limited to primary studies, including human and animal studies, published in English between 2000 and 2017. The search resulted in 12 and 22 original research articles for human and animal studies respectively, which are relevant to this topic.

Evidence from animal studies

The effects of SERMs on BMD

Fitts et al. [25] compared the efficacy of tamoxifen (1 mg/kg) and testosterone propionate (TP; at 1 mg/kg and 10 mg/kg) alone or in combination in preventing various effects of orchidectomy in adult male rats. Both tamoxifen and TP increased proximal tibia BMD of ORX rats. Rats treated with tamoxifen had a proximal tibial BMD 4% higher than testes-intact controls. The combination of tamoxifen and TP at 1 mg/kg yielded greater BMD than those achieved with tamoxifen alone [25]. In the subsequent study, Fitts et al. [26] showed that tamoxifen (0.5 mg/kg) evoked a 4.2% increase in proximal tibia BMD of rats lacking testosterone and a 9.8% increase in rats given testosterone as a co-treatment in hypothyroid (induced with 0.03% methimazole) ORX rat model. The effects of tamoxifen in protecting the bone loss in castrated male rats was replicated in ORX male strain H (Velaz Prague) rats [27]. The changes in bone density and bone mineral of tibia resulting from castration were not only entirely prevented after treated with tamoxifen (0.1 mg/day/day), but also increased above the level of intact mice after 3 months of treatment [27]. However, one study reported the negative effect of tamoxifen on bone. The study by Karimian et al. [28] showed that tamoxifen (40 mg/kg/d; p.o.) reduced the cortical BMC of the tibia in male Sprague-Dawley rats (4-week old) after 1 and 4 weeks of treatment. This was caused by a decrease in the cortical bone area without a corresponding change in volumetric BMD. It was suggested that this undesirable effect was a result of premature fusion of the growth plate and compromised final height of the bone. The decrease in cortical bone area might be due to the attenuation of periosteal and endosteal growth [28].

Broulik and Broulikova [29] showed that another SERM, raloxifene, at 5.0 mg/kg/day for three months completely prevented the decrease in BMD and BMC in ORX mice. Onoe et al. [30] showed that reduction in BMD at femoral distal metaphysis in androgen deficiency mice caused by orchidectomy was prevented with raloxifene treatment (100 µg/kg, daily) for 2 weeks. Folwarczna et al. [31] found that BMD and BMC at tibia, femur and L4 vertebra in ORX male rats treated with raloxifene were comparable with the control rats. After standardizing these values with body mass (mg/100 g of body mass), BMD and BMC of raloxifene-treated rats were significantly greater than the control rats. Besides, raloxifene also increased the BMC/bone mass ratio by 3.92% and 6.10% in the tibia and L4 vertebra, respectively [31]. The bone-protective effects of raloxifene (3 mg/kg) and risedronate (5 µg/kg) individually or in combination in osteoporotic male Wistar rats (aged 12 weeks) induced by orchidectomy have been examined [32]. After 6 weeks of treatment, results showed that administration of raloxifene and risedronate individually, and in combination caused a significant increase in BMD at the proximal and distal femur of the rats. However, finding showed that co-administration of raloxifene and risedronate rather than individually produce more favorable protection effect on bone [32]. The changes of bone quality in trabecular BMD under estradiol, raloxifene and testosterone were compared using ORX Sprague-Dawley rats (3 months old) [33]. Findings showed that raloxifene (96.35 ± 24.18%) nearly compensated the effect of castration on bone after 12 weeks of treatment. Meanwhile, bone loss at metaphysis of the tibia was only partially prevented in rats treated with estradiol (86.78 ± 11.46%) and testosterone (80.95 ± 16.7%) [33]. Another study in ORX male Sprague Dawley rats using similar treatment and duration also showed a significant enhancement in both total BMD for raloxifene and estradiol [34]. Concurrent effects of low-dose tacrolimus, an immunosuppressant, and raloxifene were also evaluated in mature male Wistar rats. Tacrolimus has been reported to increase bone resorption and bone formation at the same time, but its net effect on bone is unclear [35]. Administration of low-dose tacrolimus (0.3 mg/kg p.o. daily) and raloxifene together resulted in changes similar with those caused by raloxifene (5 mg/kg) alone, evidenced by significant increases in the bone mass/body mass ratio, BMC/body mass ratio and BMC/bone mass ratio in comparison with the control rats [36]. Although low-dose tacrolimus could improve bone health, its effects were not enhanced by raloxifene [36].

The skeletal effects of third-generation SERMs on BMD have also been studied in animal models. In a study, the effects of lasofoxifene in different doses (1, 10 and 100 µg/kg) for 60 days on BMD were determined using ORX male Sprague-Dawley rats (10-month old) [37]. Lasofoxifene at a dose of 10 or 100 µg/kg significantly prevented the decreases in total femoral BMD, distal femoral BMC, and BMD induced by aging and orchidectomy. Results found that total femoral BMD of the rats was significantly higher when treated with lasofoxifene at 100 µg/kg compared to 10 µg/kg [37]. Interestingly, a study by Sato et al. [38] showed that raloxifene (3 mg/kg, 5 times/week), bazedoxifene (3 mg/kg, 5 times/week) and tamoxifen (1 mg/kg, 5 times/week)-treated ORX wild-type C57BL/6 J mice showed BMD values similar with sham controls. Role of different domains of ER-α for the effects of estradiol and SERMs on bone mass in males were examined using an ORX male ER-α-deficient mice (20-weeks old) [39]. Three weeks after receiving the raloxifene (120 μg/mouse/day), lasofoxifene (8 μg/mouse/day), and bazedoxifene (48 μg/mouse/day), total body areal BMD showed an increment to a similar extent in ORX rats [39].

The effects of SERMs on bone microstructure and histomorphometry indices

The effects of SERMs on bone microstructure have been widely reported in animal studies. Starnes et al. [40] reported that 12 weeks of 4-hydroxytamoxifen administration increased cortical wall thickness, trabecular bone volume (BV/TV), connectivity and number as well as decreased trabecular separation (Tb.Sp) in normal male C57BL/6 J mice. Another study further supported that administration of tamoxifen (0.3 mg/kg) significantly increased cancellous bone area, Tb.Th and Tb.N, whereas Tb.Sp was significantly decreased compared with vehicle-treated ORX rats [41]. The beneficial effects of tamoxifen (40 mg/kg) supplementation have also been reported, whereby cortical periosteal circumference was improved after four week of tamoxifen supplementation in young normal male rats [42]. For the comparison of different doses of tamoxifen, an in vivo study showed that trabecular parameters evaluated by µCT remained unaffected in animals receiving two doses of tamoxifen (190 mg/kg). Treatment of four and 10 doses of tamoxifen significantly increased trabecular BV/TV, Tb.N and decreased Tb.Sp, indicating the improvement of bone microstructure in animals [42]. Zhong et al. [43] also compared the effects of four consecutive intraperitoneal administration of tamoxifen (1, 10 or 100 mg/kg/day) or single intraperitoneal administration of tamoxifen (100 or 300 mg/kg). The results indicated that four doses of tamoxifen (100 mg/kg) or a single dose of tamoxifen (100 or 300 mg/kg) increased trabecular BV/TV in the distal femur of normal young mice [43].

Raloxifene, another SERMs drug, has been demonstrated to significantly improve the trabecular network at the dosage of 3.35 mg at average/day in castrated osteoporotic male rats, thus suggesting the bone structure was preserved [34]. These observations were further supported by Onoe et al. [30] demonstrating that raloxifene restored BV/TV in ORX rats. Another animal study pointed out the comparison between the treatment of raloxifene (5 mg/kg) alone or combined with tacrolimus at two different dosages (0.3 or 0.6 mg/kg) on the osseous tissue in male rats [44]. The results showed that raloxifene alone increased width of osteoid only, whereas the combination of raloxifene and tacrolimus at 0.3 or 0.6 mg/kg increased traverse growth of bone and osteoid width as well as transverse cross-section of the cortical bone and the marrow cavity and increased thickness of trabeculae [44]. Interestingly, Folwarczna et al. [31] also found that oral administration of raloxifene hydrochloride (5 mg/kg) only increased the width of osteoid but not in all other parameters in the tibia of mature male rats. A study by Meixner et al. [45] found that combination of raloxifene (0.5 or 2.0 mg/kg) and zoledronate (a bisphosphonate, 0.06 mg/kg) for 8 weeks of treatment significantly increased BV/TV, thickness, number, cortical area and thickness in C57BL/6 J mice. Khedr et al. [32] reported that treatment with raloxifene (3 mg/kg) alone increased Tb.Th and osteocyte number of proximal tibial in ORX rats, whereas the combination of raloxifene and risedronate (5 µg/kg) decreased osteoblast and osteoclast number while increased osteocyte number. Additionally, cortical bone loss (the decrease in cortical thickness) in ORX mice was completely prevented by tamoxifen [27] and raloxifene [29].

In a study conducted by Ke et al. [37], histomorphometric analysis showed that lasofoxifene had a dose-dependent effect in preventing bone loss induced by aging and orchidectomy in the adult rats. Treatment with 1 µg/kg/day lasofoxifene for 60 days significantly decreased eroded perimeter in ORX rats, but had no effect in other parameters for bone microstructure. Orchidectomized rats treated with 10 µg/kg/day lasofoxifene had significantly higher Tb.Th as well as lower percent labeling perimeter, percent osteoid perimeter, percent eroded perimeter and bone formation rate. Orchidectomized rats treated with 100 µg/kg/day lasofoxifene had greater BV/TV, thickness, wall width, formation period, and resorption period and significantly lower labeling perimeter, osteoid perimeter, eroded perimeter, mineral apposition rate, bone formation rate/BV, bone formation rate/BS and activation frequency compared with ORX controls. These results showed that lasofoxifene prevented bone loss by inhibiting high bone turnover due to androgen deficiency [37]. In the subsequent study, Ke et al. [46] evaluated the long-term (6 months) treatment of lasofoxifene (0.01 mg/kg and 0.1 mg/kg per day) on bone histomorphometry in 15 months old male rats. Age-related increase in Tb.Sp and eroded surface in proximal tibia were completely prevented by lasofoxifene. Apart from that, the age-related decrease in BV/TV, thickness, as well as an increase in osteoclast number and surface in vertebral cancellous bone were prevented by lasofoxifene [46].

Borjesson et al. [39] compared various SERMs drugs and found that only lasofoxifene (8 μg/mouse/day) significantly increased cortical thickness in ORX male ERα-deficient mice, but raloxifene (120 μg/mouse/day) and bazedoxifene (48 μg/mouse/day) had no effect during 3 weeks of treatment. The study by Sato et al. [38], which investigated the effects of three different SERMs drugs (raloxifene, bazedoxifene and tamoxifen) on bone histomorphometric parameters, found that these drugs prevented the reductions of BV/TV, Tb.N and the elevation Tb.Sp in ORX mice after five weeks.

Despite these positive results, Karimian et al. [28] found that treatment with tamoxifen (40 mg/kg) for 4 weeks significantly decreased cortical periosteal circumference. They suggested that tamoxifen caused persistent retardation of longitudinal and cortical radial bone growth in young male rats [28].

The effects of SERMs on bone markers

In the study by Filipovic et al. [41], the effect of tamoxifen as a potent inhibitor of bone resorption on thyroid C-cells was investigated in middle-aged ORX rats. Subcutaneous injection of tamoxifen citrate at the dose of 0.3 mg/kg for 3 weeks significantly decreased serum osteocalcin (OC) level in treated rats compared to ORX rats. The level of serum calcium (Ca²⁺) and phosphate in tamoxifen-treated rats were significantly lower than the level detected in sham rats, whereas urinary Ca²⁺ was lower when compared to ORX group [41]. The findings of this study were in parallel with the study by Khedr et al. [32], which found that bone turnover markers were decreased by the treatment of SERMs. The anti-resorptive effect of these agents could prevent bone loss in androgen-deficient male rats [32]. A study by Jardi et al. [42] showed that serum OC was significantly increased following 10 days treatment of high-dose (190 mg/kg) of tamoxifen to C57BL/6 J male mice. In another study, tamoxifen at 100 mg/kg produced a significant reduction in the serum bone resorption marker, C-terminal telopeptide (CTX) in 1-month-old C577/BL6–129 male mice [43]. However, it did not change the level of bone formation markers [43].

Oral administration of raloxifene, subcutaneous injection of risedronate and their combination to the ORX rats resulted in the modulation of bone makers and suppression of bone turnover [32]. Following 6 weeks of the experimental period, both raloxifene (3 mg/kg, 3 times/week) and risedronate (5 µg/kg, 2 times/week) treatment significantly reduced alkaline phosphatase (ALP) and acid phosphatase (ACP) activities, urinary deoxypyridinoline and urinary Ca²⁺ when compared to ORX group. However, significant reductions in serum OC and serum Ca²⁺ were only found in ORX rats treated with risedronate. The combination of raloxifene and risedronate significantly reduced serum OC and ALP, urinary deoxypyridinoline and Ca²⁺, and increased Ca²⁺ content of femur ash. This study indicated the combination of raloxifene and risedronate might exert enhanced antiresorptive activities and normalize bone turnover [32].

The effect of long-term lasofoxifene treatment against age-related changes in bone mass in 15-month-old intact male Sprague-Dawley rats was evaluated in a study by Ke et al. [46]. Administration of 0.01 or 0.1 mg/kg lasofoxifene for 6 months completely prevented the age-related increase in serum OC. The serum bone resorption marker was not measured in this study but the inhibition of bone resorption by lasofoxifene was evident by the histomorphometric evaluation [46].

The effects of SERMs on bone strength

In a study by Zhong et al. [43], the effect of tamoxifen on bone strength was evaluated in 1-month-old C577/BL6–129 male mice. They were intraperitoneally injected with tamoxifen at 0, 1, 10 or 100 mg/kg for four consecutive days. Femoral 3-point bending test showed that maximal load was increased in both groups treated with 10 and 100 mg tamoxifen, but lower doses of tamoxifen did not significantly change the mechanical parameters [43]. The long-term effects of tamoxifen on bone growth and mineralization using young intact male rats was evaluated in a study by Karimian et al. [28]. Four-week treatment of 40 mg/kg of tamoxifen was found to decrease structural toughness and ultimate strength of the femur. The finding of this study was inconsistent with other studies that report the positive effect of SERMs on bone mechanical properties. The authors suggested that tamoxifen may cause persistent retardation of bone growth and might compromise bone strength in young male rats [28]. In a recent study, high-dose tamoxifen (190 mg/kg) resulted in a reduction of total cross-sectional tissue area (Tt.Ar) in in six-week-old C57BL/6 J male mice as a consequence of the reduction in medullary or marrow area (Ma.Ar) and polar moment of inertia (J) [42]. Since Ma.Ar and J are surrogates of bone strength, this may indicate compromised bone strength following high-dose tamoxifen treatment [42].

The change of bone biomechanical strength was evaluated in ORX rats receiving an average of 3.35 mg raloxifene-supplemented food per day for 12 weeks [33]. All biomechanical parameters of the metaphyseal tibia, including maximum load, failure load, stiffness and yield load value were significantly improved in ORX rats receiving raloxifene-supplemented food when compared to the group receiving soy-free pelleted food per se. From the findings, the authors suggested that raloxifene could protect cortical bone from osteoporotic fractures. Although the osteoporotic changes manifest mainly in the metaphyseal area, most of the experimental mechanical tests have been carried out in the diaphysis of the femur or tibia. The three-point bending test was only possible in the area of long bones. The author suggested that the results of breaking test allow conclusions regarding the mechanical properties of normal and osteoporotic bone [33]. Another study by Stuermer et al. [34] found that that 3.35 mg of raloxifene per day in the diet for 12 weeks significantly improved the values of yield load and maximum load in the same animal model. These findings further substantiate the protective effect of raloxifene against bone loss associated with androgen deficiency [34]. More recently, the effect of raloxifene on bone mechanical material properties was studied in C57B1/6 J male mice undergoing zoledronate treatment [45]. Low dose (0.5 mg/kg) or high dose (2.0 mg/kg) of raloxifene was given for 5 days in a week, between 20 and 24 weeks of study period. Both doses of raloxifene resulted in significantly higher mechanical properties in terms of ultimate load and displacement values compared to vehicle-treated animals. This finding suggested that, raloxifene treatment could enhance the mechanical material properties of the tissue in animals previously treated with zoledronate [45].

Prolonged use of immunosuppressive drug therapy such as methotrexate has been linked to bone loss. A study on the effect of raloxifene on the mechanical strength of the femur in male rats treated with methotrexate was done by Nowinska et al. [47]. Treatment with 5 mg/kg raloxifene significantly increased the load need to fracture the femoral neck in rats receiving methotrexate. However, the increase in the ultimate and breaking load values of the femur did not reach statistical difference [47]. In another study, the effect of raloxifene co-administered with low-dose tacrolimus on bone mechanical properties in male rats was evaluated [36]. The study found that concurrent administration of raloxifene at 5 mg/kg and tacrolimus for 4 weeks did not produce any significant effect on bone mechanical properties of the femur [36]. The evaluation of the effect of raloxifene on the skeletal system of healthy mature male rats was done by Folwarczna et al. [31]. Daily oral treatment with raloxifene hydrochloride at a dose of 5 mg/kg for 4 weeks did not significantly affect the mechanical properties of the whole femur and femoral neck. The investigation of the effects of raloxifene in a condition where sex steroid level was normal may explain the lack of significant influence of this drug on the studied parameter [31].

The effect of lasofoxifene at 1, 10 or 100 μg/kg per day for 60 days on maximal load and stiffness of the fifth lumbar vertebra (LV5) in ORX Sprague-Dawley rats were determined by compression test [37]. Lasofoxifene at 20 and 100 μg/kg increased the maximal load of LV5 but did not alter the stiffness [37]. In another study, the assessment of bone strength in aged intact male Sprague-Dawley rats treated with 0.01 mg/kg or 0.1 mg/kg lasofoxifene for 6 months showed that the decrease in mechanical strength of LV5 was completely prevented [46]. The value of ultimate strength, stiffness, toughness and elastic modulus in lasofoxifene-treated rats were significantly higher when compared to vehicle controls [46]. Mouse models lacking ERα were ORX and treated with raloxifene, lasofoxifene or bazedoxifene for the evaluation of the effects of SERMs on bone mass [39]. Findings of this study showed that none of the treatments increased bone strength in the ORX ERαAF-1-deficient mice. This study validated that the effects of SERMs on bone were contributed solely through interaction with estrogen receptor [39]. The effects of SERMs on bone health in animal studies have been summarized in Table 1.

Table 1. The effects of SERMs on bone health in animal studies.

Researcher (year)	Animal model	Treatment and dose	Duration	Mode	BMD	Micro-structure	Bone markers	Bio-mechanical
Jardi et al. (2017) [42]	Male C57BL/6J mice	Tamoxifen (190 mg/kg) – 2 doses	2 days	po	–	↔	Oc: ↔	↔
		Tamoxifen (190 mg/kg) – 4 doses	4 days		–	↑	Oc: ↔	↔
		Tamoxifen (190 mg/kg) – 10 doses	10 days		–	↑	Oc: ↑	↓
Kaczmarczyk-Sedlak et al. (2006) [44]	Mature male Wistar rats	Raloxifene (5 mg/kg, daily)	4 weeks	po	–	↔	–	–
		Raloxifene (5 mg/kg, daily) + tacrolimus (0.3 mg/kg)			–	↑	–	–
		Raloxifene (5 mg/kg, daily) + tacrolimus (0.6 mg/kg)			–	↑	–	–
Ke et al. (2000) [37]	ORX male Sprague-Dawley rats (10-month-old)	Lasofoxifene (1 µg/kg, daily)	60 days	po	↔	↔	–	↔
		Lasofoxifene (10 µg/kg, daily)			↑	↑	–	↑
		Lasofoxifene (100 µg/kg, daily)			↑	↑	–	↑
Ke et al. (2001) [46]	Male Sprague-Dawley rats (15-month-old)	Lasofoxifene (0.01 mg/kg, daily)	6 months	po	–	↑	Oc: ↓	↑
		Lasofoxifene (0.1 mg/kg, daily)			–	↑	Oc: ↓	↑
Khedr et al. (2013) [32]	ORX male Wistar rats (12-week-old)	Raloxifene (3 mg/kg, 3 times/week)	6 weeks	po	↑	↑	Oc: ↔	–
		Raloxifene (3 mg/kg, 3 times/week) + risedronate (5 µg/kg, 2 times/week)			↑	↑	Oc: ↓	–
Meixner et al. (2017) [45]	C57BL/6J male mice (16-week-old)	Raloxifene (0.5 mg/kg, 5 times/week) + zoledronate	4 weeks	s.c	–	↑	–	↑
		Raloxifene (2.0 mg/kg, 5 times/week) + zoledronate			–	↑	–	↑
Nowinska et al. (2007) [47]	Methotrexate- treated male Wistar rats	Raloxifene hydrochloride (5 mg/kg, daily)	28 days	po	–	–	–	↔
Onoe et al. (2000) [30]	ORX male	Raloxifene (100 µg/kg, daily)	2 weeks	s.c	↑	↑	–	–
Sato et al. (2017) [38]	ORX C57BL/6 male mice	Raloxifene (3 mg/kg, 5 times/week)	5 weeks	Intragastric infusion	↔	↑	–	–
		Bazedoxifen (3 mg/kg, 5 times/week)			↔	↑	–	–
		Tamoxifen (1 mg/kg, 5 times/week)			↔	↑	–	–
Borjesson et al. (2013) [39]	ORX male ERα-deficient mice (20-weeks-old)	Raloxifene (120 μg/mouse/day)	3 weeks	s.c	↑	↔	–	↔
		Lasofoxifene (8 μg/mouse/day)			↑	↑	–	↔
		Bazedoxifene (48 μg/mouse/day)			↑	↔	–	↑
Broulik (2000) [27]	Adult ORX male strain H (Velaz Prague) rats (n = 32)	Tamoxifen (0.1 mg/day)-mix in diet	3 months	po	↑	↑	–	–
Broulik and Broulikova (2007) [29]	Adult ORX male strain H (Velaz Prague) rats (n = 32)	Raloxifene (5.0 mg/kg/day)- mix in diet	3 months	po	↑	↑	–	–
Filipovic et al. (2015) [41]	ORX male Wistar rats (15-month-old) (n = 24)	Tamoxifen (0.3 mg/kg, daily)	3 weeks	s.c	–	↑	Oc: ↓	–
Fitts et al. (2001) [26]	ORX male Sprague Dawley rats (78-80 day old) (n = 48)	Estradiol (E2) benzoate (50 µg/kg) 3 times weekly	6 weeks	s.c	↑	–	–	–
		Testosterone (T3) (10 µg/kg) once daily		i.p	↔	–	–	–
		Tamoxifen (0.5 mg/kg) once daily		s.c	↔	–	–	–
		T3 + tamoxifen			↑	–	–	–
		T3+ E2 benzoate			↑	–	–	–
Fitts et al. (2004) [25]	ORX male Sprague Dawley rats (79–84 days old) (n = 44)	Testosterone 17b-propionate (TP) (1 mg/kg and 10 mg/kg)daily	6 weeks	s.c	↔	–	–	–
		Tamoxifen (1 mg/kg) daily			↔	–	–	–
		TP + tamoxifen			↑	–	–	–
Folwarczna et al. (2009) [36]	Mature male Wistar rats (n = 28)	Raloxifene hydrochloride (5 mg/kg) daily	4 weeks	po	↑	–	–	↔
		Tacrolimus + raloxifene hydrochloride			↑	–	–	↔
Folwarczna et al. (2007) [31]	Adult male Wistar rats (n = 14)	Raloxifene hydrochloride (5 mg/kg)	4 weeks	po	↑	↔	–	↔
Starnes et al. (2007) [40]	C57BL/6J mice	4-hydroxytamoxifen (0.1 mg, bi-weekly begin at 5 weeks)	12 weeks	i.p	–	↑	–	–
Stuermer et al. (2009) [34]	ORX Sprague-Dawley rats (3 months old)	Raloxifene (3.35 mg at average/day)	12 weeks	po	↑	↑	–	↑
Sturmer et al. (2006) [33]	ORX Sprague-Dawley rats (3 months old)	Raloxifene (3.35 mg at average/day)	12 weeks	po	↑	–	–	↑
Zhong et al. (2015) [43]	Wild-type male mice	Tamoxifen (1 mg/kg/day) – 4 consecutive days	1 month	i.p	–	↔	OC: ↔	↔
		Tamoxifen (10 mg/kg/day) – 4 consecutive days			–	↔	OC: ↔ CTX: ↔	↑
		Tamoxifen (100 mg/kg/day) – 4 consecutive days			–	↑	OC: ↔ CTX: ↓	↑
		Tamoxifen (100 mg/kg/day) –once			–	↑	–	–
		Tamoxifen (300 mg/kg/day) –once			–	↑	–	–
Karimian et al. (2008) [28]	Male Sprague Dawley rats (4-week-old)	Tamoxifen (40 mg/kg, daily)	4 weeks	po	↓	↓	–	↓

Evidence from human studies

In another randomized, double-blind, placebo-controlled, phase-III trial, the effects of second generation SERMs, toremifene (80 mg/kg) on BMD were evaluated in men (n = 197; aged: ≥50 years) with serum prostate-specific antigen (PSA) of 4 ng/ml or less [48]. After 12 months of treatment, there was a significant increase in BMD at lumbar spine, total hip and femoral neck by 1.6%, 0.7% and 0.2% respectively when compared with control group [48]. Interestingly, another study carried out in a larger sample population of men (n = 1284) for 2 years showed that toremifene (80 mg/kg) treatment caused a reduction in relative risk for new fracture by 50% (95% CI –1.5 to 75.0, p = 0.05) [49]. Besides, a significant increase in BMD at the lumbar spine, hip and femoral neck, as well as concomitant decrease in bone turnover markers, were observed [49]. In a similar study of toremifene (80 mg/kg) treatment in men treated with androgen deprivation therapy (ADT) (n = 647), toremifene significantly increased BMD in all measured sites after 24 months, causing a decrease in relative risk of new fracture by 79.5% (95% CI 29.8–94.0, p < 0.005) and an absolute risk reduction of 3.8% [50]. Levels of bone-specific ALP, CTX and OC were also significantly decreased when compared with the baseline which indicated low bone turnover. Besides a variation in age gap for phase-III trial of toremifene in men younger than 80 years is warranted to confirm these findings as they found that older age appears to be a potentially important risk factor for toremifene treatment [50].

Another study by Doran et al. [51] enrolled a total of 50 elderly men (mean age: 69.1 ± 6.0 years) to assess the effects of 6 months treatment with raloxifene (60 mg/day) on bone turnover markers. The results demonstrated that no significant difference was detected in the bone-specific ALP level between the raloxifene-treated and placebo groups [51]. Thirty healthy men (age: 60–70 years) were given raloxifene (120 mg/day) for 3 months to explore the effect on bone turnover marker [52]. Treatment of raloxifene significantly increased levels of luteinizing hormone (LH), follicle-stimulating hormone (FSH), total testosterone, free testosterone, but had no effect on urinary hydroxyproline/creatinine ratio as a marker for bone resorption [52]. A randomized placebo-controlled, double-blind, two-sequence crossover study consisted of 43 healthy eugonadal men (mean age: 56 years) were conducted by Uebelhart et al. [53] to investigate the effects of raloxifene on bone remodeling. In this study, the subjects were assigned with raloxifene (120 mg/day) or placebo for 6 weeks, followed by a 2-month washout period, before crossing over. The findings revealed that treatment with raloxifene increased sex hormone in serum (total testosterone, bioavailable testosterone, total estradiol, bioavailable estradiol), with a concomitant decrease in OC and ALP in serum [53]. In a recent study, a total of 159 men with advanced prostate cancer aged 45–80 years were involved to evaluate the effects of an experimental oral SERM (GTx-758) on testosterone concentrations [54]. The findings showed that GTx-758 caused reductions in free testosterone as well as bone turnover markers (CTX and bone-specific ALP) [54].

In a randomized control trial study by Smith et al. [24], men with nonmetastatic prostate cancer (n = 48) who had received a GnRH agonist treatment for 6 months or more were randomly assigned into raloxifene (60 mg/day) treatment group or nontreated group. After 12 months of treatment, BMD of the posteroanterior lumbar spine, total hip, trochanter and femoral neck of raloxifene-treated subjects were significantly increased compared to nontreated subjects. Biochemical markers of bone turnover for amino-terminal propeptide of type-I collagen and urinary excretion of deoxypyridinoline decreased by 20.3 ± 9.3% and 6.4 ± 7.8%, respectively. These findings suggest that treatment with a raloxifene is associated with a persistent lower bone turnover level and steadily increase in BMD. However, the outcome of this study may be influenced by unintended effects on subject behavior due to the open-label design [24].

The risk of femoral shaft and subtrochanteric fractures among the users of bisphosphonates and raloxifene between January 1996 and December 2006 was evaluated from a nationwide cohort study in Denmark [55]. The assessment of the crude incidence rate ratios (IRRs) for risk of subtrochanteric fractures before and after anti-osteoporotic treatment was performed. The study showed an increased fracture risk after the start of several bisphosphonates therapy [alendronate (IRR =2.41, 95% CI 1.78–3.27), clodronate (IRR= 20.0, 95% CI 1.94–205) and etidronate (IRR =1.96, 95% CI 1.62–2.36)] but no increased risk was observed after the start of raloxifene (IRR =1.06, 95% CI 0.34–3.32). However, an increased risk of subtrochanteric fracture was also present before the initiation of the bisphosphonate and raloxifene therapy. These data indicate that the increased femoral and subtrochanteric fractures may be confounded by the underlying disease being treated [55]. The effects of SERMs on bone health in human studies have been summarized in Table 2.

Table 2. The effects of SERMs on bone health in human studies.

Researcher (year)	Study type	Characteristics of the subjects	Interventions (Drug and dose)	Study period	Findings
Smith et al. (2004) [24]	Randomized controlled trial	Men with nonmetastatic prostate cancer receiving a GnRH agonist for 6 months or more (n = 48)	Raloxifene (60 mg/day	12 months	Men receiving raloxifene had higher total hip and trochanter BMD and lower bone markers (amino-terminal propeptide of type I collagen by and urinary excretion of deoxypyridinoline)
Smith et al. (2008) [48]	Randomized, double-blind, placebo controlled, phase III trial	Men with prostate cancer and serum PSA ≤4 ng/ml (n = 197; age: ≥50 years)	Toremifene (80 mg/day)	12 months	Toremifene increased BMD at lumbar spine, total hip and femoral neck.
Smith et al. (2010) [49]	Randomized, double-blind, placebo controlled phase III trial	Men receiving ADT for prostate cancer and at increased risk for fracture or osteopenia at the femoral neck or L1–L4 lumbar spine (n = 1284; age: ≥50 years)	Toremifene (80 mg/day)	2 years	Toremifene reduced relative risk of new vertebral fractures, increased BMD at lumbar spine, total hip, and femoral neck Toremifene decreased bone turnover markers (ALP, C-telopeptide and OC)
Smith et al. (2011) [50]	Randomized, double-blind, placebo controlled, phase III trial	Men with prostate cancer and serum PSA ≤4 ng/ml (n = 647; age: ≥50 years)	Toremifene (80 mg/day)	24 months	Toremifene decreased relative risk of new vertebral fracture and incidence of non-traumatic fracture Toremifene increased BMD at lumbar spine, total hip and femoral neck. Toremifene decreased bone-specific markers (ALP, C-telopeptide and OC)
Doran et al. (2001) [51]	Randomized blinded study	Elderly men (n = 50; age: ≥60 years)	Rolaxifene (60 mg/day)	6 months	Level of bone-specific ALP did not differ between the raloxifene-treated and placebo groups.
Duschek et al. (2004) [52]	Double-blind, placebo-controlled study	Healthy men (n = 30; age: 60–70 years)	Raloxifene (120 mg/day)	3 months	Raloxifene increased LH, FSH and sex steroid hormones No effects on bone resorption marker.
Uebelhart et al. (2004) [53]	Randomized placebo-controlled, double-blind, two-sequence crossover study	Healthy eugonadal men (n = 43; age: 49–70 years)	Raloxifene (120 mg/day)	6 weeks raloxifene (8 weeks washout period)	Raloxifene increased total testosterone, bioavailable testosterone, total estradiol, and bioavailable estradiol levels Raloxifene decreased serum OC and ALP.
Yu et al. (2015) [54]	Prospective randomized 1:1:1 phase 2 open-label multicenter trial	Men with advanced prostate cancer (n = 159; age: 45–80 years)	GTx-758 1000 mg (two 500 mg tablets) or 2000 mg	300 days	GTx-758 decreased free testosterone, CTX, and bone-specific ALP.
Lin et al. (2014) [77]	Cohort study	Osteoporosis patients with newly diagnosed vertebral or hip fracture (n= 68,808; age: >50-year old)	Raloxifene	Between 2003 and 2007	There was no excess risk for osteonecrosis of the jaw in osteoporosis patients aged >50 years old using raloxifene.
Vestergaard et al. (2011) [55]	Nationwide cohort study (Denmark)	Patients exposed to drugs against osteoporosis	Raloxifene	Between1996 and 2006	Increased risk of subtrochanteric fractures was present before the start of raloxifene but not after the initiation of the therapy.

Discussion

Previous studies on the use of SERMs in animal models of male osteoporosis generally yielded positive findings, indicated by improved BMD, BMC, bone histomorphometric indices and microstructures. The effects of SERMs on bone biomechanical strength were marginal because it would require a longer time for any changes to occur. Few studies reported harmful skeletal effects of SERMs in young animals, probably due to inhibition of normal bone growth. The findings from human studies generally suggested SERMs were beneficial in increasing BMD at various sites, particularly at lumbar spine, total hip and femoral neck. This was achieved by suppressing bone resorption, indicated by reduced serum bone turnover markers. Studies on the effects of SERMs on fracture risk are limited [49,50]. Clinical trial on fracture risk would require larger samples size and longer study period [56]. Besides, all the human studies on SERMs and bone health in men were conducted in North American and European regions. It is not known whether ethnic difference in the response to the drug exists. Furthermore, more studies on the comparison in efficacy between SERMs and other osteoporosis medications should be conducted to validate their use in men at risk for osteoporosis.

Endogenous 17β-estradiol is a skeletal protective hormone in both women and men [57]. Estrogen replacement therapy was clinically used in preventing bone loss in postmenopausal women but its use has declined due to long-term side effects such as increased in breast cancer and cardiovascular events [58–61]. Low-dose ERT, though has been associated with some benefits in some trials [62,63], is not used clinically in men. Selective estrogen receptor modulators exhibit unique ability as an agonistic and antagonistic to ER depending on which tissues of their activity are examined. This would circumvent the adverse effects of ERT in female and feminizing effect in men. Currently, there are two main chemical classes of SERMs approved for clinical used, which are tamoxifen (first generation) and raloxifene (second generation) [64,65]. The newer generation of SERMs such as lasofoxifene and badezoxifene also have been introduced [37,66].

The skeletal effects of SERMs are exerted through ER. Both ERα and ERβ have been detected in cultured human osteoblast-like cells, and in uncultured osteoblast cells [67–70]. Selective estrogen receptor modulators diffuse into the cell and bind to the ERs, triggering dimerization of ER and regulation of target gene expression [71]. Tamoxifen has been shown to inhibit bone resorption in neonatal rat osteoclast cultures at micromolar concentrations [72]. Raloxifene was shown to induce osteoclast apoptosis, and increase mineralization and bone nodule formation in bone marrow culture [73]. Raloxifene also increased the osteoblast-specific transcription factor Cbfa1/Runx2 and alpha2 procollagen type-I chain mRNA. It also inhibited the release of pro-inflammatory cytokines (interleukin-6 and tumor necrosis factor-α) by osteoblast [74]. Additionally, raloxifene enhanced the production of osteoprotegerin by osteoblast [75] and decreased the responsiveness of osteoclasts to receptor activator of nuclear factor kappa B ligand (RANKL) and macrophage-colony stimulating factor (M-CSF) [76]. Overall, SERMs suppress osteoclast formation and activity while promoting the osteogenesis. Since the predominant effect of SERMs is suppression bone resorption, they are classified as antiresorptive agents.

Selective estrogen receptor modulators have several advantages over other osteoporosis medications. Osteonecrosis of the jaw (ONJ) is a concern among patients using antiresorptive therapy. A retrospective cohort study was carried out to compare the risk of ONJ between oral alendronate (bisphosphonates) and raloxifene/calcitonin (nonbisphosphonates) in the Taiwanese population [77]. This study has found that osteoporotic patients receiving raloxifene/calcitonin treatment have no excess risk of developing ONJ. In another cohort study, the risk of ONJ in men and women Taiwanese osteoporotic patients taking oral alendronate was compared with a group of patients taking raloxifene only. From the findings, only one out of total eligible patients using raloxifene was diagnosed as having ONJ. However, only women subjects were analyzed in this study, as raloxifene was not licensed to treat male osteoporosis.

Selective estrogen receptor modulators also exert extra-skeletal benefits to men with prostate cancer. In a study conducted by Yu et al. [54], researcher assigned GTx-758 (an oral SERMs agonist) or leuprolide (a GnRH agonist) to men with advanced prostate cancer. The study indicated that leuprolide reduced total testosterone in a greater proportion as compared to GTx-758, whereas GTx-758 was superior in lowering free testosterone and PSA. Recently, SERMs were reported to have the potential to treat breast cancer, infertility and idiopathic gynecomastia in men. For instance, tamoxifen has been proposed as a treatment to prevent gynecomastia in prostate cancer patient receiving hormonal ablation, specifically the nonsteroidal anti-androgens [78]. Another study on the survival benefit of tamoxifen and aromatase inhibitor among male and female breast cancer patients has revealed that overall survival of tamoxifen-treated patient is better than in patients receiving aromatase inhibitor treatment [79]. The overall 5-year survival for tamoxifen-treated female and male breast cancer patients was similar, 85.1% and 89.2%, respectively, whereas aromatase inhibitor-treated patients had 73.3% of 5-year overall survival.

In terms of side-effects, a review on randomized controlled trials (RCT) and non-RCT studies that include the safety data of tamoxifen treatments in men showed that gastrointestinal and cardiovascular problems were the most commonly reported adverse events in prostate cancer patients [80]. For most men, tamoxifen is well tolerated, but psychiatric side effects (decreased libido, anxiety and sleep disorder) were reported among male breast cancer patients. Since tamoxifen has become an off-label option for the treatment of male breast cancer, infertility and idiopathic gynecomastia, patients should be well informed about the risk although the reported adverse events are very minimal. Based on the available evidence, studies that rigorously document the safety profile of SERMs in men are still lacking.

Conclusions

As a conclusion, the documented evidence in this review suggested the potential of SERMs on male skeletal system. However, the use SERMs should be under careful consideration due to the reported side effects. More studies are required for a better understanding of the mechanism of action and its advantages over anti-resorptive agents for the treatment of osteoporosis.

Disclosure statement

The authors declared no conflict of interest.

Additional information

Funding

We thank Universiti Kebangsaan Malaysia for funding the studies via [GUP-2017–012] and [GUP-2017–060] grants.