An emerging role in bone pathophysiology has been attributed to the immune system, giving rise to a new field of research termed osteoimmunology. The concept of osteoimmunology refers to the mutual interactions between the immune system and bone. In the English literature the term “osteoimmunology” was coined in 2000 by Arron and Choi (2000[6]); nevertheless, one of the authors of this review article (PP) already had used the German term “Osteoimmunologie” in 1997.

At the cellular level, the osteoclast, the cell responsible for bone resorption, can be regarded as the prototype of an osteoimmune cell: osteoclasts share common precursor cells with monocytes, macrophages, and (myeloid) dendritic cells. Accordingly, first insights into the osteoimmunological crosstalk were gained by studies on interactions of immune cells and osteoclasts leading to bone destruction in inflammatory diseases such as periodontitis or rheumatoid arthritis (Horton et al., 1972[61]). By now, it has become clear that cells of the immune and bone system have many molecules, such as transcription factors, signaling factors, cytokines, or chemokines, in common (Tsukasaki and Takayanagi, 2019[124]). One of the first type of immune cells that have been shown to mediate effects of the immune system on bone are T cells, in particular CD4⁺ cells. In rheumatoid arthritis, a typical osteoimmune disorder characterized by bone erosions in multiple joints in conjunction with inflammation of the synovium, stimulation of bone resorption by osteoclast is exclusively mediated by the Th17 subset of CD4⁺ cells (Kotake et al., 1999[71]; Sato et al., 2006[110]). These cells accumulate in the synovial fluid of patients with RA and promote osteoclastogenesis in the first place by the secretion of IL-17 that stimulates RANKL expression by syno-vial fibroblasts (Hirota et al., 2007[59]; Kontake et al., 1999[71]; Nistala et al., 2008[86]; Sato et al., 2006[110]). Additionally, IL-17 augments local inflammation. Hence, the production of inflammatory cytokines such as TNF and Il-6 is increased, which in turn amplifies RANKL expression and thereby indirectly fuels osteoclastogenesis (Sato et al., 2006[110]). In recent years, the most potent pro-osteoclastogenic CD4⁺ subset was further narrowed down to a particular type of Th17 cells, which derives from FOXP3⁺ T cells. They have been shown to loose expression of the transcription factor FOXP3 under arthritic conditions, thereby promoting the development of Th17 instead of regulatory T-cells (T_reg) (Komatsu et al., 2014[70]).

To sum up, tremendous progress has been made in elucidating protagonists and pathways contributing to bone destruction in conjunction with inflammatory diseases, in particular with rheumatoid arthritis (Tsukasaki and Takayanagi, 2019[124]). The contribution of the immune system to bone loss seen in osteoporosis, in contrast, is less well understood. Cessation of ovarian function and, associated therewith, loss of estrogen has been known since nearly eight decades to be a key event in promoting accelerated bone loss in early menopause (Albright et al., 1941[3]), and, as proposed by the unitary model for the pathophysiology of primary or involutional osteoporosis, also in slow bone loss in late post menopause, referred to as age-related bone loss, and in elderly men (Riggs et al., 1998[104]). Effects of estrogen loss are to some extend mediated by a direct modulation of osteoblast, osteoclast, and osteocyte physiology via estrogen receptors on these cells. Specifically, estrogen loss increases the number of osteoclasts and at the same time decreases the number of osteoblasts leading to an unbalanced activity of the basic multicellular unit in favor of bone resorption (Clarke and Khosla, 2010[28]). However, it is now well accepted, that effects of estrogen deficiency, in particular on bone resorption, are mainly indirect via the release of bone-active cytokines. Among these osteoclastogenic cytokines are inflammatory cytokines, suggesting a role of interactions of the immune system and bone tissue also in the pathophysiology of osteoporosis. A role of proinflammatory cytokines in osteoporosis is strengthened by the findings of elevated levels of TNF, IL-1, IL-6, or IL-17 in the first ten years after menopause (Pacifici et al., 1990[90]) and in osteoporotic postmenopausal women compared to non-osteoporotic postmenopausal women (Zhao et al., 2016[135]). Moreover, the phenomena of “inflammaging” characterized by a low-grade inflammatory status and increased levels of proinflammatory markers in elderly people (Franceschi et al., 2000[48]) has in several studies been linked to bone loss and fracture risk (Cauley et al., 2007[22]; Ding et al., 2008[42]; Ganesan et al., 2005[51]; Nakamura et al., 2011[82]; Pasco et al., 2006[95]). Likewise, chronic inflammatory diseases, e.g. rheumatoid arthritis or Crohn's disease, promote an osteoporotic phenotype (Blaschke et al., 2018[15]; Sapir-Koren and Livshits, 2017[109]).

As in rheumatoid arthritis, also in osteoporosis T-cells are thought to be a major source of proinflammatory cytokines (Pietschmann et al., 2016[98]; Rauner et al., 2007[102]). Our group has shown that the proportion of CD8⁺ cells that express TNF is expanded in postmenopausal women with osteoporotic fractures when compared to age-matched controls (Pietschmann et al., 2001[97]). D'Amelio et al. described a higher production of TNF in T-cells and monocytes (D'Amelio et al., 2008[37]) and Zhao et al. an upregulated expression of IL-17 in CD4⁺ cells of osteoporotic postmenopausal women (Zhao et al., 2016[135]). However, an important link between estrogen loss, T-cell-dependent inflammation, and osteoporosis has been unraveled only recently. Cline-Smith et al. for the first time described a molecular mechanism by which estrogen loss promotes a low-grade inflammation by T-cells during the acute phase of bone loss in ovariectomized mice (Cline-Smith et al., 2020[29]). They provide data suggesting a central role of bone marrow dendritic cells (BMDCs) in the induction of proinflammatory cytokines producing T-cells. Ovariectomy (OVX) in their study increases the number of BMDCs and subsequently the amount of IL-7 and IL-15 produced by these cells. IL-7 and IL-15, in turn, induces antigen-independent production of IL-17A and TNF in a subset of memory T cells (T_MEM). Further, a crucial role of the suggested pathway in OVX associated bone loss was confirmed by showing that T-cell-specific ablation of IL15RA prevented IL-17A and TNF expression and an increase in bone resorption or bone loss after OVX (Cline-Smith et al., 2020[29]). Interestingly, the authors speculate that a crucial role of the activation of T_MEM in postmenopausal bone loss may provide an explanation for varying susceptibilities to develop osteoporosis within a population. Varying levels of T_MEM in individuals reflect different lifetime exposures to commensal and pathogenic microbes and might be associated with higher or lower levels of TNFα and IL-17A and in consequence of bone loss induced after menopause (Cline-Smith et al., 2020[29]).

Another subclass of T-cells increasingly evidenced to act at the interface of the immune and skeletal system are regulatory T (T_reg) cells. They are characterized by the expression of the transcription factor FOXP3 and their main function is to suppress multiple types of immune cells and prevent excessive immune reactions, inflammation, and tissue damage. Their role in bone biology is clearly anti-osteoclastogenic (Bozec and Zaiss, 2017[20]). Accordingly, the transfer of T_reg cells into T-cell deficient mice was associated with an increased bone mass and a decreased number of osteoclasts (Zaiss et al., 2010[134]). Moreover, Foxp3 transgenic mice were protected from OVX-induced bone loss, supporting a role of T_reg cells in bone loss in conjunction with estrogen deprivation (Zaiss et al., 2010[134]) Consistent with this, estrogen has been shown to stimulate proliferation and differentiation of T_reg cells (Tai et al., 2008[117]).

A role of B-cells in the pathophysiology of osteoporosis is supported by the finding that they produce RANKL and OPG, and, hence, act as regulators of the RANK/ RANKL/OPG axis (Walsh and Choi, 2014[127]). Indeed, production of RANKL by B-cells is increased in postmenopausal women (Eghbali-Fatourechi et al., 2003[45]) and B-cell ablation of RANKL in mice partially protects from trabecular bone loss after ovariectomy (Onal et al., 2012[88]). A role of B-cells in bone metabolism and osteoporosis is further strengthened by the results of a global gene expression study by Pineda et al.. Comparing gene expression in OVX mice and control mice they identified several pathways attributed to B-cell biology among the top canonical pathways affected (Pineda et al., 2014[99]). A more recent study compared global gene expression in B-cells obtained from the bone marrow of OVX and control mice (Panach et al., 2017[92]). In a second stage, they studied the association of polymorphisms in selected differentially expressed genes in postmenopausal women and identified a significant association of single nucleotide polymorphisms (SNPs) in CD80 with bone mineral density (BMD) and the risk of osteoporosis. A possible link between this molecule and BMD might be indirect via its costimulatory function for the activation of T-cells or direct via the described inhibitory effect on osteoclast generation (Bozec et al., 2014[21]). To sum up, substantial evidence for a contribution of B-cells to the development of osteoporosis exists. However, the exact mechanism linking estrogen deficiency to B-cells and bone loss seen in postmenopausal women remains incompletely understood.

A novel and rapidly expanding field deals with the influence of the gut microbiome (GM) on a person's health and provides exciting new insights into the crosstalk between the homeostasis of bone metabolism and the intestinal flora (Behera et al., 2020[10]; Ding et al., 2020[43]; Pacifici, 2018[90]). It is now well accepted that the GM, the entirety of microorganism living in the human digestive tract, influences development and homeostasis of gastrointestinal (GI) tract tissues and also of tissues at extra-GI sites (e.g nutrient production and absorption, host growth, immune homeostasis). Moreover, complex diseases such as type 1 and 2 diabetes, transient ischemic attack, or rheumatoid arthritis have been linked to changes in the composition of the GM (Behera et al., 2020[10]). Sjogren et al. have shown that germ-free mice exhibit increased bone mass and thereby first evidenced a relation between bone homeostasis and the GM (Sjogren et al., 2012[113]). Additional support for this crosstalk comes from experimental data showing that modulation of the GM by the use of probiotics or antibiotics affects bone health (Guss et al., 2019[57]; Li et al., 2016[73]; Ohlsson et al., 2014[87]; Parvaneh et al., 2015[94]; Rozenberg et al., 2016[106]). An important evidence for a role of the GM in estrogen driven bone loss comes from a study showing that germ-free mice are protected from trabecular bone loss induced by sex steroid deprivation (Li et al., 2016[73]).

Various mechanisms have been proposed to modulate this close “microbiota-skeletal” axis, one of them being the effects of the GM on host metabolism. The GM has been shown to influence the absorption of nutrients required for skeletal development such as calcium, and thereby affect bone mineral density (Rodrigues et al., 2012[105]). Absorption of nutrients might be influenced by intestinal pH values, which depend on the composition of the GM. Additionally, microbial fermentation of dietary fibers to short chain fatty acids (SCFAs) seems to play an important role in this process. In adults, consumption of different prebiotic diets that can be fermented to SCFAs was associated with an increased resorption of calcium (Whisner et al., 2014[129], 2016[130]). Beyond this influence on intestinal nutrient absorption, SCFAs have emerged as potent regulators of osteoclast differentiation and activity and of bone metabolism (Zaiss et al., 2019[133]). For instance, in mice fed with SCFAs or a high-fiber-diet an increase in bone mass was observed. Moreover, postmenopausal as well as inflammation-induced bone loss was prevented and the protective effect was associated with impaired osteoclast differentiation and bone resorption (Lucas et al., 2018[77]). SCFAs are therefore an example of gut-derived microbial metabolites that diffuse into the systemic circulation. By doing so, these substances can regulate anatomically distant organs such as the skeletal system (Zaiss et al., 2019[133]).

A well-established function of the GM is to modulate immune functions. Hence, effects of the GM on intestinal and systemic immune responses, which in turn modulate bone homeostasis, provide another important link between the GM and the skeletal system. Bone active cytokines released by immune cells in the gut or immune cells activated in the gut and then circulating to the bone are discussed as mechanisms most likely mediating this GM-immune-bone axis (Pacifici, 2018[90]). Among cells of the immune system, Th17 cells and T_reg cells are thought to play a prominent role in this crosstalk. Specifically, the balance of Th17/T_reg cells has been shown to be modulated by gut macrobiotics (Dar et al., 2018[39][40]), and also here a key role in promoting the differentiation and proliferation of T_reg cells is attributed to SCFAs (Arpaia et al., 2013[5]; Furusawa et al., 2013[49]; Smith et al., 2013[114]; Zaiss et al., 2019[133]).

Another unexpected link between SCFAs, the immune system, and bone metabolism was reported only recently. Li et al. demonstrated that the bone formation stimulating effect of intermittent parathyroid hormone (PTH) treatment depends on SCFAs, in particular butyrate, produced by the microbiome (Li et al., 2020[74]). Further, they provided evidence for butyrate, in concert with PTH, to induce CD4⁺ T cells to differentiate into T_reg cells, which in turn stimulate CD8⁺ T cells to produce Wnt10b (Li et al., 2020[74]). Wnt10b is a key activator of Wnt signaling in stromal cells and osteoblasts and is known to promote bone formation by increasing osteoblast proliferation, differentiation, and survival (Monroe et al., 2012[80]). Also PTH induced bone loss has been known to depend on T cell activation (Gao et al., 2008[53]; Tawfeek et al., 2010[118]), but it was not clear if these T-cells originate in the bone marrow or in the intestine. Only recently, it was shown by Yu et al. that bone loss induced by PTH depends on activation of intestinal TNF⁺ and Th17 T cells in response to the gut microbiota and recruitment of these cells to the bone marrow (Yu et al., 2020[132]). Taken together, increasing evidence attributes the microbiome and metabolites produced by the microbiome, in particular SCFAs, a key regulatory function in bone homeostasis. Probiotics and interventions targeting the GM and its metabolites might therefore be a promising future strategy for the prevention and treatment of osteoporosis.

Cellular senescence describes a cell fate induced by various types of stress and is associated with irreversible cell cycle arrest and resistance to apoptosis (Hayflick, 1965). In addition, cells entering senescence are characterized by excessive production of proinflammatory cytokines, chemokines, and extracellular matrix-degrading proteins, a state referred to as senescence-associated secretory phenotype (SASP) (Tchkonia et al., 2013[120]). The number of senescent cells increases during the process of aging (Tchkonia et al., 2010[119]), which has been evidenced to play a major role in age-related tissue dysfunction and the development of several age-related diseases such as diabetes mellitus, hypertension, atherosclerosis, or osteoporosis (Khosla et al., 2020[68]).

An important contribution to understanding the role of senescence in the development of osteoporosis was made by Farr et al. only a few years ago. They have shown that B cells and T cells, myeloid cells, osteoprogenitors, osteoblasts, and osteocytes are among the types of cells within the bone microenvironment becoming senescent with aging, and that an increased production of key SASP factors with aging can be seen particularly in senescent myeloid cells and osteocytes (Farr et al., 2016[46]). Further, they provided evidence for an accumulation of senescent cells in bone biopsy samples from older postmenopausal women compared to younger premenopausal women (Farr et al., 2016[46]). In a more recent study, Farr et al. provided a causative link between cellular senescence and age-related bone loss. They demonstrated that elimination of senescent cells or inhibition of their SASP prevented age-related bone loss in mice by decreasing trabecular and cortical bone resorption and increasing or maintaining bone formation on endocortical and trabecular surfaces, respectively (Farr et al., 2017[47]). Further, consistent with the finding of an increased SASP particularly in osteocytes (Farr et al., 2016[46]; Piemontese et al., 2017[96]), the observed bone sparing effect was mediated partly by the elimination of senescent osteocytes (Farr et al., 2017[47]). To sum up, considerable progress has been made in understanding the role and mechanisms of senescence in age-related bone loss. Based on these new findings, targeting cellular senescence by senolytics and senostatics has emerged as a potential promising therapeutic strategy for the treatment of age-related osteoporosis.

The goals of osteoporosis therapy are to reduce fracture risk and bone loss, prevent disability, and control pain (Spencer, 1982[115]). Prevention strategies include fall reduction, correcting impaired vision or hearing, reducing fall risk inducing drugs (FRIDs), establishing muscle training, balance training, quit cigarette smoking and alcohol intake, and adequate intake of vitamin D, protein and calcium (Akkawi and Zmerly, 2018[1]). An impressive number of evidence has been published in the last years with enormous impact on clinical application. Therefore, it seems reasonable to continue this narrative review on current treatment options (for an overview see Table 1(Tab. 1); References in Table 1: Bischoff-Ferrari et al., 2009[12]; Black and Rosen, 2016[14]; Chesnut et al., 2004[27]; Compston et al., 2017[30]; Cosman et al., 2016[33]; Cummings et al., 2009[35]; D'Amelio and Isaia, 2013[38]; Kanis et al., 2013[65]; Neer et al., 2001[84]; Reid et al., 2009[103]).

The treatment for osteoporosis in the first step consists of basic interventions such as prescribing exercise, weight-bearing physical activity and exercises that improve balance and posture, a diet rich in vitamin D and calcium, quitting smoking, and limiting alcohol use, measures summarized under point “Nonpharmacological treatment” In the second step, specific medication prescriptions are necessary (Figure 1(Fig. 1)). Medication groups are divided in anti-resorptive (anti-catabolic) drugs, anabolic drugs, and combinations of remedies (Ukon et al., 2019[125]). Osteoclasts remain the main targets of medical intervention, even though osteoblasts emerge as new cells of interest in that topic. The Wnt signaling pathway has become a fascinating spot for new medical interventions; Dickkopf-1 and sclerostin, both inhibitors of this pathway, are promising therapeutic targets (Chen et al., 2019[25]).

Nevertheless, only a minor part of eligible patients is treated with any of these therapeutic options. In a population-based study published by Lorentzon et al. only a proportion of 21,8 % received an adequate treatment (Lorentzon et al., 2019[76]). Non-adherence is particularly to blame for the undertreatment and therefore of interest. Main reflections on adherence are mentioned under the heading “Adherence”.

A protective effect against hip and non-vertebral fractures can be achieved with vitamin D doses ≥ 800 international units (IU) daily in combination with a calcium intake between 700 and 1200 mg/day (preferable by dietary intake). This dose of vitamin D, effective in reducing the risk of falls, is recommended in women and men ≥ 50 years, and can be increased in patients with higher risk of fractures. However, the intermittent prescription of large doses of vitamin D (≥ 100.000 IU) is associated with an increased risk of fractures and falls (Sanders et al., 2010[108]). The combination of calcium and vitamin D is necessary in patients treated with bone specific therapies (to achieve the full effects shown in the intervention trials) and to prevent secondary hyperparathyroidism, hypomagnesemia, and disturbances of bone metabolism (Bolland et al., 2015[16]; Compston et al., 2017[30]). Vitamin D is fat-soluble; for optimal resorption it is important to provide it with the meal.

Based on clinical algorithms published in guidelines (Figure 3(Fig. 3); References in Figure 3: Kanis et al., 2019[64]; Cosman, 2020[32]; Anastasilakis et al., 2020[4]), specific medication must be offered to patients with a high risk of fractures or patients with a fragility fracture (Kanis et al., 2019[64]; Qaseem et al., 2017[101]). No study has been powered to show differences in risk reduction of fractures between the different treatment modalities. Nevertheless, osteoanabolic drugs reduce the risk of vertebral and nonvertebral fractures in a faster mode than antiresorptive drugs and should be considered as first line therapy in special clinical circumstances such as in patients with prior fragility fractures, multiple fractures during the clinical course (Cosman, 2020[32]), and a very low bone mineral density (t-score below - 3) (Cosman, 2020[32]). In general, the choice of treatment is made on clinical judgement considering potential side effects, effects on different skeletal parts, and costs.