Is an association of acro-osteolysis, bone fragility, and enchondromatosis a newfound disease caused by an amplification of PTHLH? A case report

Acro-osteolysis (AO) is a clinical symptom and a radiological sign characterized by the destruction of the distal phalanx of the fingers and toes. Two types are described: (i) resorption of the distal end of the phalanx, and (ii) damage to the body of the phalanx (band or transverse AO). The second type is characterized by the presence of a hypodense line through the body of the distal phalanx by standard radiography.

Clinically, AO presents initially as a “drumstick” deformation of the fingers. As the bone breaks up, the deformation gradually worsens, giving the fingers and toes a shortened, soft, enlarged appearance. Deformations are sometimes associated with other clinical signs (short stature, deafness, skin abnormalities, neuropathy, etc.) related to the etiology.

There are many causes of AO, both constitutional and acquired [1]. Acquired AO is usually asymmetrical and may affect only one phalanx. Environmental causes are dominated by repetitive microtrauma resulting only in affected fingers. Toxic substances such as vinyl chloride or snake venom can cause more diffuse forms but not symmetrical damage. Phalangeal tumours are singly located and associated with typical radiological abnormalities.

In the context of non-constitutional general diseases, acquired AO is only an associated sign, in most cases secondary to the original disease. The cause may be neurological (diabetic neuropathy), vascular (systemic lupus erythematosus, scleroderma, vasculitis with poly-angiitis), inflammatory (psoriasis, psoriatic rheumatism) or infectious (e.g. leprosy).

Hyperparathyroidism can also lead to AO, usually more diffuse and generally associated with a broader context of bone demineralization, which may speciously suggest a genetic origin. Phosphocalcic exploration is also essential for establishing one AO etiology.

Constitutional forms are unusual, but recognized (e.g. in some lysosomal diseases). These diseases are very rare and associate AO with other clinical signs. These signs guide diagnosis: they can be bone manifestations (fragility, deformities, malformations, worm-like bones), skin signs (pachydermia, hyperhydrosis, lipodystrophy), neurological signs (loss of sensitivity, neuropathy, muscle weakness) or haematological signs [2‐15]. The best known is the Hajdu-Cheney syndrome linked to a NOTCH2 mutation [16, 17]. Although these syndromes are numerous, they affect only a few patients. None of them associates AO and multiple enchondromas. A summary of clinical and genetic features is given in Table 1.

A 15-year-old girl of Caucasian French origin was referred to a paediatric rheumatologist for recurrent pain in her fingertips in the last few months, with gradual onset. She had no past medical history, and no surgical history except for one tonsillectomy. She was born with a monozygous twin to non-consanguineous parents.

At first visit, she complained of pain in her fingers and toes that had appeared a few months previously. Progressively, she presented drumstick fingers, and then multiple bone pain localization (metatarsals, wrists).

Pulmonary function test, cardiac investigation and capillaroscopy were normal. Sensory or motor impairment, joint or skin involvement, calcinosis, Raynaud’s syndrome, signs of dysthyroidism or diabetes were not found. Respiratory examination was normal, and the differential diagnosis of hypertrophying osteoarthropathy could be excluded.

otological examination found an auditory defect at 45 dB only at 2000 Hz. Ophthalmological examination reported mild hyperopia.

She presented some clinical features not reported in other members of her family: slender frame, high forehead, facial asymmetry, moderate scoliosis. Her growth was regular and homogeneous.

The patient declared no repeated microtrauma of extremities or any unusual exposure to vinyl chloride.

AO is a radiological and clinical manifestation of an underlying disease that may or may not be of constitutional origin.

This duplication partially overlapped a polymorphism reported in the database of genomics variants, hindering interpretation. Besides this polymorphism, five genes remain: REP15, MRPS35, MANSC4, KLHDC and the PTHLH gene. To date, in the literature, duplications 12p11.22-p11.23 have been described in five families. The first patient described by Collinson et al. presented short stature and very unusual generalized symmetrical enchondromatosis of the tubular bones [20]. Three patients described by Gray et al. presented AO, cortical irregularity of long bones and metadiaphyseal enchondromatosis [18]. Recently Limenis et al. published a review on differential diagnosis of AO and its distinguishing features, with particular attention paid to types more commonly encountered in paediatrics [21]. Our article reports new findings on the relationship between this amplification and this disease expressed by an AO and an anomaly of the long bones. It also provides data on anatomopathology and clinical evolution. Flöttmann et al. reported on a three-generation pedigree with short humerus, curved radius, and a specific type of severe brachydactyly with features of types E and A1 but without the enchondromatosis and AO hitherto described [22]. Finally, Tacke et al. reported on a patient with multiple skeletal abnormalities including chondrodysplasia, lesions radiographically resembling enchondromas and posterior rib deformities leading to a severe chest deformity [23]. No AO was observed in this patient (probably not yet present owing to the young age of the patient compared to the other observations).

A summary of the phenotypic characteristics of our patient and those described in the literature is given in Table 2.

PTHLH is a humoral factor known to regulate the balance between proliferation and differentiation of chondrocytes during endochondral bone development. It is produced by the perichondrial and chondrocyte cells, which grow during skeletal development, and it maintains these cells in a proliferative state by a paracrine mechanism. It has a large structural homology with parathyroid hormone (PTH). Mutations and deletions of the PTHLH or PTHR1 genes (encoding the PTHLH protein receptor) cause several types of skeletal dysplasia, including brachydactyly type E, Eiken syndrome, Jansen’s metaphysical chondrodysplasia, and Blomstrand’s chondrodysplasia. Given the role of the PTHLH protein, it is likely that a variation in the number of copies of this gene is involved in the patient’s bone disease. Studies in mice showed that overexpression of PTHLH was responsible for a form of dwarfism with short limbs and delayed endochondral ossification [24]. The 12p11.22 replication most recently described by Flöttmann et al. is also the smallest reported to date (about 70 kb) and contains only the PTHLH gene [22]. Although it appears to cause skeletal signs comparable to those described in patients described by Gray et al. [18] and Collinson et al. [20] (short humerus, radial abnormalities, brachydactylia), AO associated with bone cysts is not present in patients with this small duplication. Importantly, the duplications described in these two articles, of larger size, respectively 851 kb and 502 kb, in addition to the PTHLH gene and besides the polymorphism previously mentioned, contained three additional genes: REP15, MRPS35 and KLHDC. MRPS35 encodes a mitochondrial ribosomal protein. REP15 is involved in the recycling of the transferrin receptor. The KLHDC gene belongs to the Kelch-like gene family, particularly expressed for limbs in mice during the embryonic period. All these data suggest that the isolated duplication of the PTHLH gene alone may not suffice to generate an AO process associated with enchondromatosis, but is the likely cause of other skeletal growth abnormalities identified in the literature and consistent with the results of studies in mice. It could be argued that the duplication of genes included in the larger genomic imbalances, such as those described by Gray et al. and Collinson et al., or in our study, could also affect the bone phenotype, modulating the expression of PTHLH or exerting a synergistic effect. Finally, it is possible that these duplications 12p11.22-p11.23, including the PTHLH gene, reflect redesigns of varying phenotypic expressivity that explain the intra-family and inter-family phenotypic differences. For example, our patient’s mother carried the same amplification but had no clinical or radiological sign of AO or any enchondromatosis. She had no disease or comorbidity, except obesity. One possible explanation for this phenotypic variability is the “multiple hit”, whereby a second insult is needed to produce a more severe clinical phenotype, via an additive or synergistic effect on other skeletal growth pathways. The second hit might be a disruptive single base-pair mutation in a phenotypically related gene or even an environmental event. Concerning the monozygotic twin sister, we could not legally perform a test because she was asymptomatic. This suggests variable penetrance of the disease, or the involvement of regulatory genes. The reproducibility of symptoms to PTHLH amplification in the different case reports suggests a causal relationship.

A query of the Database of Chromosomal Imbalance and Phenotype in Humans using Ensemble Resources (DECIPHER; https://​decipher.​sanger.​ac.​uk) identified 11 patients, with detailed phenotype descriptions, presenting with a 12p amplification encompassing at least PTHLH. Skeletal anomalies were identified in 6/11 patients and features overlapping our patient’s in 3/6 patients with skeletal anomalies (308,811, 385,237 and 295,002). We note that the variation identified overlaps a small duplication, only including the PTHLH gene, in patient 308,811 associated with radial bowing, short humerus and type A1 brachydactyly and most likely corresponding to the duplication described by Flöttmann et al. [22].

To date, 12p11.22-p11.23 duplications have been reported in five families with skeletal abnormalities, and in particular AO and enchondromatosis associated with bone fragility. This new observation, associated with the few cases recently reported in the literature, suggests a close relationship between the presence of this microduplication and the skeletal abnormalities found in the patient, so strengthening the hypothesis of a new nosological entity with a close relationship with disruption of the PTHLH-PTHR1 signalling pathway during skeletal development. The clinical features seem to be variable, possibly by regulation sequences included in the duplication area, and more explorations are needed to understand the mechanisms. We suggest the descriptive name ABES (acro-osteolysis, bone fragility and enchondromatosis syndrome) to designate this disorder.