Perspective on skeletal health in inflammatory bowel disease

Inflammatory bowel disease (IBD) is characterized by a systemic, chronic, relapsing, and disbalanced inflammatory response, primarily originating from and expressed in the intestinal mucosa. IBD mainly comprises two distinct phenotypes: Crohn’s disease and ulcerative colitis.

The incidence of IBD is worldwide rising with a north-to-south gradient and appears to be introduced in the non-Western world concurrent with industrialization, and the accompanying dietary habits [1]. In the Netherlands, the annual incidence increased from 17.90 per 100,000 in 1991 to 40.36 per 100,000 in 2010 for IBD; from 5.84 per 100,000 to 17.49 per 100,000 for Crohn’s disease; and from 11.67 per 100,000 to 21.47 per 100,000 for UC [2]. The onset typically occurs in the second and third decade of life, yet may start at any age. IBD is most commonly characterized by periods of active disease (relapse) and periods of remission (quiescent) [3]. IBD diagnosed in children and in adulthood is associated with osteopenia, osteoporosis, and fracture risk in several studies, both in Western [4‐8] and in Asian [9] countries. The increased relative risk for osteoporosis or bone fractures when compared with a normal or general population is still a matter of debate. In a recent systematic review and meta-analysis, based on nine studies comparing IBD patients with general controls, the relative risk for global fracture was 1.38 (95% CI 1.11–1.73). For vertebral fractures, odds ratio was 2.26 (95% CI 1.04–4.90). In this paper, mean BMD and Z-scores for IBD patients versus controls were decreased at all sites, and patients with IBD had an increased risk of fractures [10]. Vertebral fracture risk was also reported to be elevated in another meta-analysis considering IBD patients [11].

Skeletal health comprises a well-balanced bone metabolism necessary for growth and maintenance of bone tissue in order to function adequately in the skeleton. Healthy bone tissue protects from bone fracture due to low-energy, mechanical trauma. A biomarker of bone health is bone density, measured by dual-energy X-ray absorptiometry (DXA). Diagnosis of impaired bone health in IBD patients primarily depends on DXA measurements expressed in T-scores, sometimes Z-scores, whereas bone turnover markers in serum or urine are not being used in standard clinical practice. Similarly, bone biopsy and quantitative histopathology or CT scanning is unusual, but may be done in a research setting. Assessment of osteoporosis in younger patients has been extensively reviewed elsewhere [12].

As an indication of the burden caused by a diminished skeletal health in IBD patients, a prevalence of osteopenia from 32 to 36% and a prevalence of osteoporosis from 7 to 15% have been reported [13, 14]. The estimated elevated excess risk for a fragility fracture in Crohn’s disease patients compared with the general population was 30 to 40% more fractures [15, 16]. An elevated risk of fractures has also been reported in patients suffering from ulcerative colitis, particularly at initial diagnosis [5]. Together with the data from an extensive review and meta-analysis comparing this burden with general controls, it seems advisable to identify patients with IBD who are at increased risk for osteoporosis in order to prevent fractures [10].

The etiology of bone loss in IBD is multifactorial (see Table 1). It includes the following: (I) variables primarily associated with skeletal health in general, such as potentially insufficient intake of calcium, vitamin D, magnesium, and potassium; smoking; decreased sun exposure; a low peak bone mass; a low body mass index; and decreased physical activity.

Additionally, an unbalanced diet and a dysbiotic intestinal microbiome, probably interacting, and partly uncharacterized and non-specific, may contribute to impaired bone health [17]. (II) IBD-associated factors

Most of the above-mentioned factors are—partly—uncharacteristic or atypical for IBD but occur concurrent with IBD.

Therapy is ideally primarily based on evidence-based medicine. Thus, convincing data from statistically well-powered trials in relevant populations are crucial. Unfortunately, these trials are sparse, particularly when considering robust clinical end points, i.e. reduction of bone fractures. Arguments for choice of treatment of impaired bone health are therefore also based on pathophysiological and clinical reasoning. To improve bone health, simple measures to change life style, diet, or medication may be helpful. Nutritional supplements, discontinuation of smoking, avoidance of (long-term) use of detrimental drugs, and increase of physical activity are well worth advising in general. Treatment specifically aimed at IBD-associated causes of decreased bone health is less well characterized. Rational, pathophysiology-based therapy needs further understanding of the (patho)physiological mechanisms underlying this IBD-associated osteoporosis. We therefore elaborate on the current knowledge regarding systemic inflammatory disease, such as in IBD, and the associated changes in bone health.

In several studies, it has been shown that elevated concentrations of circulating pro-inflammatory cytokines including tumor necrosis factor alpha (TNF-α), interleukin (IL)-1β, IL-6, and IL-17 are present in IBD patients. Crohn’s disease is reported to be associated with elevated concentrations of the Th1 pro-inflammatory cytokines IL-2, IL-17, interferon-γ (IFNγ), and TNF-α, while ulcerative colitis is associated with Th2-profile cytokines, such as IL-4, IL-5, and IL-13 [19]. Subsequently, it has been put forward that Th1, Th2, and Th17 cells are broadly involved in the pathogenesis of IBD through regulation of inflammatory cytokine networks. All of these pro-inflammatory factors are potentially detrimental for bone metabolism and may be responsible for bone loss and increased fracture risk. In most in vitro research, effects of a single cytokine on healthy osteoclasts have been analyzed. Additionally, it has been demonstrated that growth potential and differentiation capacity of primary bone cells harvested from patients with quiescent Crohn’s disease are diminished. Yet, the response of bone cells from Crohn’s disease patients to cytokines has been reported to be similar [20]. Cytokines such as TNF-α, IL-1β, and IL-6 are known to increase the production of receptor activator of nuclear factor kappa-B ligand (RANKL) by pre-osteoblasts, thereby promoting osteoclastogenesis [21‐23]. These cytokines bind to their respective receptors, which causes increased activation of p38 mitogen–activated protein kinase (MAPK), nuclear factor kappa-B (NF-κB), and c-Jun-N terminal kinase (JNK) pathways resulting in excessive bone loss and reducing BMD. Moreover, TNF-α inhibits the activation of R-spondin 2 (RSPO2), important for osteoblast maturation, causing decreased bone mass [24].

A role of IL-1beta has been proposed in IBD patients with a IL-1β-511*2 polymorphism (carriers of allele 2 of the AvaI polymorphism), characterized by IL-1β hypersecretion, which induces a higher risk of diminished BMD than in healthy controls [25]. IL-17A, the defining cytokine of Th17 cells, has been shown to stimulate osteoclast formation [26]. However, in this study, osteoclast formation was predominantly from a non-classical blood monocyte precursor, while IL-17A did not affect the classical blood monocytes [27]. Moreover, Th17 cells express higher RANKL than other Th subsets suggesting they may indirectly promote osteoclastogenesis [28].

IL-12, a T cell mediator, is also an important inflammatory cytokine observed in IBD. It was reported that IL-12 induces apoptosis in bone marrow cells treated with TNF-α via an interaction between TNF-α-induced Fas and IL-12-induced Fas ligand (FasL), and that, as a result, osteoclastogenesis is inhibited [29, 30].

Interestingly, the effects of IL-23 on bone appear to be pleotropic. Originally, it was suggested that IL-23 may act via IL-17 to induce osteoclastogenesis [31]. But later studies showed IL-23 to upregulate RANK expression on osteoclast precursor cells [32], and to promote osteoclastogenesis in co-culture systems with osteoblasts [33]. Nevertheless, both suggested pathways lead to stimulation of bone resorption.

The effect of T cell–derived anti-inflammatory cytokines on bone cells is also extensively studied. Both interferon-γ (IFNγ) and IL-4, the signature Th1 and Th2 cytokines, inhibit osteoclastogenesis [34, 35]. IL-4 and the closely related IL-13 induced osteoprotogerin (OPG) expression in endothelial cells, which is an inhibitor of osteoclastic resorption [36]. In addition, inhibition of mRNA for RANK and nuclear factor of activated T cells 2 (NFAT2), a transcription factor, has also been noted. Consequently, IL-4 and IL-13 decrease the number of tartrate-resistant acid phosphatase–positive multinucleated cells and the mRNA expression of calcitonin receptor, tartrate-resistant acid phosphatase, and cathepsin K in mouse spleen cells and bone marrow macrophages treated with macrophage colony–stimulating factor and RANKL [37]. The effects of IL-4 and IL-13 have also been studied in osteoblasts, in which they stimulate production of IL-6 [38, 39]. Moreover, in an osteoblastic cell line (MG-63), IL-4 and IL-13 inhibit mRNA expression of B1 and B2 kinin receptors induced by either IL-1β or TNF-α, which is suggestive that these cytokines exert their anti-inflammatory effects, including impairment of bone resorption [40].

Although the predominantly negative effects of separate cytokines for bone are evident, in clinical situation, the patients have to deal with a cocktail of cytokines and inflammatory factors leaking into the circulation and thus eventually preparing the stage in bone tissue (see Fig. 1). Indeed, in Crohn’s disease patients, osteoclastogenesis has been proven to be increased via T cell mediation [20]. In only a few studies, effects of a (complete) composition of inflammatory factors have been tested by using serum from Crohn’s disease patients in an in vitro setting. In this balance-of-cytokine studies, it has been observed that bone formation measured by assaying calcium content and dry weight was decreased in intact bone explants when adding serum derived from Crohn’s disease patients, even if these patients were in remission. With light microscopy of bone tissue, a discontinuous, uneven mineralized bone matrix and disorganized osteoblasts, characterized by altered morphology, were identified when treated with this mixture Crohn’s disease patients’ derived serum [41]. This effect was counteracted by blockade of IL-6. Even in children, osteoblast function and mineralization were affected by Crohn’s disease patients’derived serum in vitro, suggesting a mechanism by which Crohn’s disease may affect bone formation and, when occurring at young age, peak bone mass. In this latter study, IL-6-neutralizing antibody did not change the negative effect on osteoblasts [42]. Data from our own lab confirmed that serum from IBD patients in both active and quiescent state of disease contained an altered composition of inflammatory factors compared to healthy control serum. Bone cells showed reduced cell proliferation after exposure to IBD patient’s serum [43]. To summarize, at the bone tissue level, cytokine profiles ascribed to the Th1 signature appear to promote osteoclastogenic bone resorption, whereas Th2-signature profiles may counteract this. This balance between Th1-Th2 profiles may finally determine outcome on bone health. Balance-of-cytokine studies are warranted to further elucidate this concept.

Lately, gut-derived serotonin, produced by enterochromaffin cells of the duodenum, has attracted attention. It was shown in mice that high serotonin concentrations in blood were associated with osteoporosis, due to decreased osteoblast proliferation via activation of 5-HT_1B receptors on pre-osteoblasts. In line with this, both increased expression of the tryptophan hydroxylase inhibitor Lrp5 (low-density lipoprotein receptor–related protein 5) and decreased tryptophan hydroxylase 1 (Tph1) expression lead to decreased serotonin generation and secondarily to an increment of bone mass [44].

In colitis, elevated concentrations of serotonin have been demonstrated. As high, gut-derived serotonin concentrations have been associated with osteoporosis, colitis may contribute to osteoporosis in IBD patients. Conversely, inhibition of gut-derived serotonin might become a treatment option for osteoporosis. These findings are now experimentally observed in animal studies, and corroboration of these findings in humans is awaited [45, 46].

Recently, an immunomodulating effect of vitamin D on both the innate and the adaptive immune system has been established. In IBD patients, vitamin D might influence disease course [47]. Vitamin D levels in IBD patients are usually low, due to multiple reasons such as decreased intestinal absorption following surgery, deficient diet, and decreased sunlight exposure. At sufficient oral intake, calcium is predominantly absorbed via a non-saturable, paracellular diffusive pathway in the duodenum and proximal jejunum, regulated via the dynamics of the epithelial tight junctions. 1,25 (OH)2D3 (calcitriol) may enhance paracellular Ca²⁺ permeability. Vitamin D absorption is dependent upon its dissolution in micelles and occurs by simple passive diffusion and partly via the paraintestinal lymph drainage system [48‐51]. In addition, sun exposure of the skin with ultraviolet B (UVB)–rich light induces vitamin D₃ (cholecalciferol) generation from 7-dihydrocholesterol, a less efficient process with aging. With respect to bone, low levels of vitamin D are correlated with secondary hyperparathyroidism [52]. Secondary hyperparathyroidism is characterized by high bone turnover leading to low bone mass and increased fracture incidence. However, in a mouse model for UC, high vitamin D was shown to have adverse effects on bone health, likely because of the synergy with circulating pro-inflammatory mediators on a resorptive effect, leading to loss of BMD [53]. This phenomenon has not yet been confirmed in a clinical setting.

A more recent therapeutic agent for treatment of osteoporosis is denosumab, a RANKL inhibitor. In IBD patients, this compound has not yet been scrutinized. This may be due to a theoretical disadvantage. RANKL inhibition by denosumab might affect NF-κB function which is pivotal in inflammation, and thus in chronic intestinal inflammation such as IBD. Interestingly, the use of denosumab seems neither related to a higher risk for inflammatory activity (relapse of IBD), nor for an increased risk for infectious diseases [79]. This has been subject of study in rheumatoid arthritis but not in IBD [80], and effects on bone formation more than on bone resorption have been proposed when observing elevated concentrations of bone-specific alkaline phosphatase [81]. In addition, denosumab use seems to be associated with a rebound effect following discontinuation of therapy which may be considered to be a major flaw when using this drug in the relative young IBD population [82]. Parathormone has anabolic effects on bone metabolism, like the recombinant hormone teriparatide. This may be an interesting medication in low peak bone mass as is found in IBD patients. However, it has been studied in corticosteroid-associated osteoporosis, but not in IBD patients as such [83].

Overall, if opting for drugs like these, robust scientific evidence is lacking and these options should be restricted to individuals with peculiar clinical circumstances, by preference in centers of excellence.

In conclusion, assessment of impaired bone health in IBD patients relies on DXA in clinical practice, with sometimes additional radiographs of the spine, whether primarily for suspected osteoporosis or secondarily as an incidental finding on abdominal CT scan (for IBD purposes). Assessment of the intestinal inflammatory burden and bone-influencing drugs like corticosteroids or TNF inhibitors form part of determination of bone health risk. Advice to introduce general bone health–promoting procedures and strict control of disease activity contribute to optimal bone health. When treating impaired bone health in IBD patients, focus should be on general advices such as calcium supplementation, sufficient vitamin D use, and life style and diet measures. Medications to improve bone health ideally should be chosen based on pathophysiological reasoning corroborated by relatively sparse specific trials in the population at risk. Timing of drugs remains a subject of study, as well as the sequence in which drugs, and in which drug classes, may be advised. More recently introduced agents are insufficiently documented for regular use in IBD patients with bone health at risk.