Cherubism as a systemic skeletal disease: evidence from an aggressive case

Cherubism (OMIM #118400) is a rare autosomal dominant genetic condition caused by mutations in the SH3BP2 gene, encoding for an adaptator protein, SH3BP2 [1]. Progressive painless bilateral enlargement of the jawbones characterizes the disease. During the growing phase, the jawbone is replaced by a granuloma containing multinucleated giant cells within a fibrous stroma [2]. Cherubism is described as a maxillofacial localized disease, only affecting jaw bones, and follows a natural course of expansion (from 2 years old to puberty), stabilization and regression. However, its penetrance and expressivity are widely variable from non-symptomatic to lethal cases [3, 4]. In addition to the maxillofacial involvement, general manifestations may occur, especially respiratory disorders mostly due to obstructive apnea [5].

Recent studies carried out on mouse models, have uncovered crucial molecular and cellular aspects of the pathogenic mechanism of cherubism, highlighting auto-inflammatory aspects of the disease and the bone involvement [6]. Notably, cherubism mice exhibit global bone loss and macrophage infiltration of internal organs [7]. However, these observations have never been reported in human cherubism, although interestingly, some authors had partially explored the bone and inflammatory markers, in order to exclude differential diagnosis, such as parathyroid impairment [2, 8‐11]. From these studies, we know that standard blood count, serum electrolytes, serum calcium and phosphate concentrations, and TSH, FSH, LH, PTH, PTHrP, T4 and T3 hormones, calcitonin and osteocalcin levels are all within the normal range [8‐10], and that alkaline phosphatase is increased [2, 9‐11]. However, systemic exploration of cherubism patient is extremely rare and incomplete. Since cherubism is caused by mutations of a ubiquitous adaptator protein involved in bone remodelling and inflammation and the knock-in mice show a systemic bone loss, we investigated potential systemic manifestations of cherubism in a young patient. Here we report a case of aggressive cherubism case associated with systemic inflammatory and bone loss features.

About 400 cases of cherubism have been reported up to date in various ethnic groups worldwide. Severity and clinical expressivity of the disease range from no clinically detectable features to grotesquely deforming jaws leading to life threating conditions [3, 20, 21], as in the present case. This case study reports for the first time a cherubism patient with systemic manifestations, and hence questions the exclusive maxillofacial tropism of this pathology as described to date [22].

Recent genetic, molecular and cellular studies have provided critical insights into the pathogenic mechanisms of cherubism thanks in large part to the creation of knock-in mouse models expressing SH3BP2 mutations responsible for the human pathology [7, 23]. However, unlike human cherubism patient who are heterozygous for the SH3BP2 mutations, heterozygous mice do not exhibit any cherubism phenotype. In contrast, homozygous mutant mice develop severe systemic bone loss related to osteoclast hyperactivity as well as systemic inflammation due to hyperactive macrophages secreting elevated levels of TNF-α [7]. Despite this new knowledge, no investigations have yet explored systemic inflammatory and bone phenotypes in human patients, presuming the disease to be exclusively maxillofacial.

The diagnosis of aggressive cherubism in our patient was made according to the standard genetic, clinical and morphopathological analyses (Fig. 1). The patient exhibits an extremely severe case of cherubism with strong nuclear NFATc1 staining as previously described [16].

Furthermore, in this case report, we showed that the patient exhibits a systemic bone loss in addition to the maxillofacial involvement. The patient presented low bone density on DXA examination (whole body aBMD Z-score − 2.8 SD) and high bone turnover, as evidenced by elevated levels of CTx, a marker of bone resorption, and P1NP and osteocalcin, markers of bone formation [24]. We eliminated other potential aetiologies by blood tests and genetic studies (hemopathy, other chronic inflammatory diseases, immune deficiencies, gene panel sequencing and nutritional deficiency). Osteoporosis in chronic inflammatory disease has been described previously, especially in paediatric inflammatory bowel disease or chronic inflammatory conditions [25]. The principal involved mechanism is a chronic systemic inflammatory process, which activates osteoclastogenesis via circulating cytokines [26]. In the cherubism mouse model, bone loss is thought to be due to both chronic inflammation and direct stimulation of osteoclastogenic differentiation [7]. The involved mechanisms might be the same as those in the pathogenesis of human cherubism.

The patient’s low bone mass was characterized by whole body and lumbar spine aBMD Z-scores (Table 2). However, the total left hip aBMD Z-score did not show low bone mass. It has already been reported that measurement at the hip is not reliable in growing children [27]. Of note, the patient is Gabonese and her Z-scores were adjusted according to stature-for-age and sex. Concerning ethnic considerations, as there are no data for African populations, we used the LUNAR ethnicity adjustment for the black ethnic group. Thus, we cannot exclude a possible bias in the interpretation of the aBMD Z-score in this patient.

Simultaneously, the patient presented a homogeneous, painless hepatomegaly without hepatitits insufficiency, associated with splenomegaly. Because of her ethnic origin, parasitological, virology, and hematological examinations were performed to exclude tropical causes of hepatomegaly and splenomegaly.

Moreover, we detected an increase of interleukin-1 and TNF-α levels in the patient’s serum.

TNF-α expression has been shown in jaw lesions in human cherubism, but TNF-α serum level has not been reported [28]. In human cherubism, it remains unclear whether increased TNF-α production is primarily induced locally in jawbone lesions by macrophages, or by circulating myeloid cells. In our case, jawbone lesions began at the age of 3, and became rapidly more aggressive at the age of 6, with systemic bone loss and inflammation. This timeline highlights that jawbone lesions may have preceded systemic manifestations. However, TNF-α alone is not sufficient to induce osteoclast differentiation and activity, but synergistically potentiates with other cytokines, such as RANKL, IL-1 and TGF-β [29].

Most of the SH3BP2 mutations causing cherubism are in the exon 9 of the gene, in a region coding for a 6 amino acid region (RSPPDG) between the PH and SH2 domains of the protein. Levaot et al. in 2011 [30] identified the Tankyrases (TNKS) as the proteins binding to this sequence. In addition, they demonstrated that TNKS by ribosylating Sh3bp2 allows its interaction with RNF146. The latter is then responsible for the ubiquitylation of Sh3bp2 hence its degradation [30]. Using the cherubism mouse model, Levaot et al. demonstrated that the Sh3bp2 mutation increased the stability of the protein, leading to its cytoplasmic accumulation [31]. Then, the increased Sh3bp2 quantity is able to enhance osteoclastogenesis and induce systemic bone loss as in our patient. However, it is not clear how this Sh3bp2 cytoplasmic accumulation in the osteoclasts primarily triggers a jaw osteolysis and only arising in deciduous or mixed dentition.

Myeloid Differentiation Factor 88 (MYD) is an adaptator protein that binds Toll-Like Receptor (TLR) and IL1R, to activate downstream transcription factors leading to cytokine production. TLR recognize pathogen-molecular patterns (PAMPs) from microorganism and damage associated molecular patterns (DAMPs) released by stressed cells or damaged tissues. In Sh3bp2^ki/ki mice, Yoshitaka et al. demonstrated that TLR2–4-MYD88 mediated pathway in hematopoietic cells was critical for cherubism phenotype development and that IL1R played a minor role in cherubism inflammation [32]. They showed that stimulation of TLR by PAMPs induced an Y183 Sh3bp2 phosphorylation by SYK, leading to an activation of NFκB pathway and a production of TNF-α in mutant macrophages and that SYK was primordial in the cherubism pathogenesis [32]. By creating germ-free Sh3bp2^ki/ki mice, they showed that bacterial stimulation was not the only trigger for cherubism and that DAMPs participated only partially to the activation of cherubism [32]. From this study, the authors proposed an explanation to the craniofacial localization of the human cherubism phenotype and the natural course of the disease. They suggested that human cherubism (as heterozygous mice) is a latent disease, activated by the detection of PAMPs from the microbial flora in the oral cavity or DAMPs secreted during teeth eruption or mechanical stress of mastication. This theory would explain the craniofacial localization of the cherubism in human, and more specifically, the DAMPs involvement during teeth eruption theory might explain the age of expression of the disease. Indeed, upon microbial stimuli on cherubism affected patient, TLR activation could initiate the process responsible for cherubism, and in severe cases, lead to systemic manifestations, by a vicious circle, as observed in autoinflammatory diseases.

In a previous study, we showed that multinucleated giant cells from granuloma are CD68-positive cells and hypothesized that these cells may differentiate into macrophages in non-aggressive cherubism and into osteoclasts in aggressive cherubism, stimulated by the NFATc1 pathway [15]. This same imbalance might be involved in the osteoporotic phenotype observed here. However more thorough explorations should be conducted before any definitive conclusions are drawn. The human cherubism trigger has yet to be identified as our previous study suggests a more secondary role for TNF-α in the human pathogenesis [15], whereas TNF-α had been largely incriminated in the mouse cherubism phenotype [33].

We may hypothesize from this case that cherubism is a systemic bone disease with a systemic skeletal phenotype that could follow the evolutionary time course of the jaw lesions. To confirm this hypothesis, a study of the osseous and inflammatory phenotypes in a larger series of cherubism patients should be carried out. We suggest also paying more attention to a potential cherubism systemic skeletal phenotype during patient management. If this systemic skeletal cherubism phenotype is confirmed, it would simplify treatment of severe cherubism patients and allay reservations about applying a systemic treatment such as those recently published (tacrolimus [34] or imatinib [35]), to a previously believed jaw-localised disease.