Introduction

Fibrodysplasia ossificans progressiva (FOP), also known as myositis ossificans progressiva, Münchmeyer’s disease and ‘stone man syndrome’, is an extremely rare autosomal dominant genetic disease involving ACVR1, which codes for the activin receptor-like kinase 2 receptor (ALK2) in people.^1,2 It results in dysregulation of the bone morphogenic protein (BMP) pathways within the connective tissues.^1,2 Pathophysiology of the disease starts with minor soft-tissue injuries that trigger an inflammatory response leading to muscle injury and necrosis, and subsequent fibroproliferation, maturing to endochondral bone formation. Regional swelling and great toe malformation of (bilateral hallux valgus) are also usually present in people.^1,2 Maxillofacial presentations affecting the masticatory muscles alone or together with extra-skeletal lesions elsewhere in the patient, resulting in unresolving trismus, have been reported in humans.^3456–7 No medical or surgical treatment has yet been proven to give any prophylactic or curative effect. Corticosteroids, non-steroidal anti-inflammatory drugs, retinoids and bisphosphonates have been used for treatment but seem to be of limited efficacy.¹ FOP carries a poor prognosis.^1,2

FOP-like conditions affecting various areas of the spine, trunk and legs have been previously reported in 11 cats.^{89101112131415–16} Terms such as generalised myositis ossificans, fibrodysplasia ossificans, progressive ossifying myositis, FOP and FOP-like disease have been used as alternatives to describe the disease. However, in only one reported feline case has similar genetic abnormality to that described in humans been identified.¹⁴ The term FOP should be reserved for those cases where the genetic abnormality has been identified; otherwise, the term ‘FOP-like’ should be used. The term ‘myositis’ should be restricted to non-hereditary forms of heterotopic ossification such as myositis ossificans circumpscripta.¹⁶ To our knowledge, this clinical case is the first report in the veterinary literature of a FOP-like condition involving the masticatory muscles.

Case description

An 8-year-old tortoiseshell spayed female domestic shorthair cat was referred for anorexia, weight loss and progressive trismus of at least 1 month duration. A history of chronic rhinitis since the age of 4 months was reported. A PCR test for calicivirus was negative. Previous antibiotic therapies (doxycyline 10 mg/kg PO q24h for 10 days; benzylpenicillin procaine 7 mg/kg [dihydrostreptomycin 10 mg/kg long-acting formulation] IM; and clindamycin hydrochloride 11 mg/kg PO q24h for 10 days) did not result in any clinical improvement. The cat was exclusively fed veterinary dry food and no history of trauma was reported.

Upon presentation, the cat had generalised weakness, bilateral purulent nasal discharge, with concurrent reduced nasal airflow, right epiphora with chemosis and third eyelid protrusion, and concurrent mild bilateral mandibular lymph node enlargement. Body condition score was 3/9 (weight 2.9 kg [6.4 lb]). Neurological examination revealed decreased bilateral palpebral reflexes on the ophthalmic branch territories of the trigeminal nerves, and an absence of pupillary light reflexes. Vision testing revealed complete visual deficits, previously unnoticed by the owners, absent menace response in both eyes, and the patient bumping into objects. Spinal reflexes were normal and no gait abnormality was present. Oral examination showed painful mouth opening with a restricted vertical mandibular range of motion (vmROM), and an inability to close the mouth (Figure 1). The rest of the physical examination was within normal limits.

Figure 1

(a) Image of an 8-year-old tortoiseshell spayed female domestic shorthair cat on initial presentation with open bite. (b) Rostral aspect of the oral cavity under general anaesthesia. Note the bilateral submucosal haematomas involving the retromolar areas (white arrows). The vertical mandibular range of motion (vmROM) was 22 mm

Serum biochemistry and electrolytes were unremarkable. Serological screening for infectious agents (Toxoplasma species, Ehrlichia species, Cryptococcus species, Anaplasma species) was negative. A complete blood count and blood smear examination revealed a mild microcytic non-regenerative normochromic anaemia with concurrent moderate neutrophilia and monocytosis. Antinuclear antibody testing was negative, and creatine kinase activity, activated partial thromboplastin time, activated clotting time, thrombin clotting times and D-dimers were within the normal limits. The restriction of vmROM persisted under general anaesthesia (22 mm). Intra-oral examination and dental charting revealed oropharyngeal swelling, with haematomas involving the retromolar areas, ulceration of the dorsal aspect of the tongue and mild generalised periodontitis.

In order to explore the cause of reduced mouth opening, CT of the head was performed. Selected reconstructed images are provided (Figure 2). CT images showed bilateral, asymmetrical, non-enhancing (49 Hounsfield units in the pre- and post-contrast images), ill-defined hypodense areas within the masticatory muscles, mostly the temporalis and masseter muscles. Irregular periosteal reaction was noted on the ramus of the mandible and on the parietal and temporal bones bilaterally. Some new bone foci extended into the temporalis and medial pterygoid muscles away from the bone surfaces, indicating an ectopic mineralisation disorder. CT findings suggested bilateral chronic masticatory myositis.

Figure 2

CT of the head. Transverse reconstructions at the level of both rami. (a) Irregular periosteal new bone on both mandibulae and parietal bones (white arrows); dystrophic mineralisation in temporalis/pterygoid muscles (black arrows). (b,c) Bilateral, non-enhancing, ill-defined hypodense areas in temporalis and medial pterygoid muscles (white arrowheads)

Six ultrasound-guided needle-core biopsies from the right and left masticatory muscles were collected and submitted to two different laboratories for comparative anatomopathological analysis. Biopsied material was submitted fresh and fixed in 10% neutral buffered formalin and decalcified as required. Fresh biopsies were isopentane pre-cooled in liquid nitrogen for cryohistology. Cryosections were stained with haematoxylin and eosin, Engel’s modified Gömöri trichrome stain, oil red O and periodic acid–Schiff reaction. Fixed biopsies were embedded in paraffin, sectioned at 3–5 µm thickness, mounted on glass slides and stained with haematoxylin and eosin before light microscopy evaluation. The samples showed extensive replacement of the myofibers with a dense vascularised mesenchymal tissue exhibiting progressive ossification. The mesenchymal tissue was composed of peripheral fascicles of activated fibroblasts and intermediate osteoblasts depositing woven bone and gradually forming trabeculae of lamellar bone (Figure 3). Rare atypia and mitoses were present in the fibroblastic population. Residual muscle fibres embedded in the lesions were atrophic and degenerated. Some samples revealed striated muscle containing occasional inflammatory infiltrates. No infectious agent was observed within the samples. A FOP-like disease was considered the most likely diagnosis by both laboratories.

Figure 3

Photomicrograph of a section of the needle-core biopsy of the right temporalis muscle. (a) Extensive replacement of the muscle (asterisks) by a dense mesenchymal tissue exhibiting progressive ossification (arrows). Haematoxylin and eosin, bar = 250 µm. (b) Proliferation of fibroblasts (white arrows) and osteoblasts (black arrow) gradually forming trabeculae of mineralised lamellar bone. Note residual atrophic muscle fibres embedded in the lesion (black arrowheads). Haematoxylin and eosin, bar = 100 µm

Supportive treatment during hospitalisation consisted of fluid therapy (3 ml/kg/h), nutrition through an oesophagostomy tube, antibiotic therapy (clindamycin hydrochloride 11 mg/kg PO q24h for 7 days, then doxycycline 10 mg/kg PO q24h), pain management (buprenorphine 20 µg/kg IV q6h), a muscle relaxant (dantrolene-sulfate 1 mg/kg PO q8h) and immunosuppressive doses of methylprednisolone (3.3 mg/kg PO q24h) in case of an immune-mediated process. The latter was subsequently replaced with robenacoxib (1 mg/kg PO q24h) 4 weeks after initial presentation.

Despite weight stabilisation, and improved willingness to move, allowing transient home care, the cat failed to recover the ability to eat spontaneously. A full-body CT scan was eventually performed 4 weeks after initial presentation and excluded additional areas of soft tissue calcification, pulmonary lesions and lymph node enlargement. Five weeks after initial presentation there had been a rapid decrease of vmROM leading to a total inability to open and close the mouth (13 mm open bite). A persistent bilateral purulent nasal discharge with almost total blockage of the nares led to the decision to perform a tracheostomy and nasal cavity lavage under general anaesthesia. During induction of anaesthesia, an emergency endotracheal intubation by tracheotomy, followed by cardiac resuscitation, had to be carried out. Because of the worsening prognosis, the owners elected humane euthanasia and rejected post-mortem investigations.

Two remaining samples of fresh biopsied material, kept at –80°C, were ultimately sent in dry-ice medium to a specialised laboratory for molecular analysis. Genomic DNA was extracted from the muscle samples, using sodium dodecyl sulfate/proteinase K lysis followed by purification on silica columns. The primers were synthesised at a 25 nmol scale, desalted and designed on the reference sequence of Genbank NC_018730, with the use of a software. After PCR amplification and product purification Sanger sequencing was performed for the whole nine coding exons of Felis catus ACVR1 (total linear DNA length 222,790,142 bp), located on the C1 chromosome, and purified. DNA sequencing products were analysed on a 96 capillary DNA analyser. Two deletions were highlighted in the exon 8 and exon 9 PCR products, and were further amplified and sequenced. Both of them were characterised in intron 8, upstream of the last coding exon 9 of ACVR1. Four bases (TTGA) were deleted around –123 bases of the exon 9 start (position 139,512 of the total gene sequence), and five bases (CCCAT) were deleted around –225 bases upstream of the exon 9 start (position 139,410 of the total gene sequence).

Discussion

This is the first case report of FOP in a cat, which selectively involved the masticatory muscles with sparing of other muscle groups. A review of previous descriptions in this species are summarised in Table 1.

Table 1

Summary of the clinical features in 11 cats diagnosed with a fibrodysplasia ossificans progressiva-like condition

Compared with previously reported cases aged from 16 weeks to 6 years, this patient was older (8 years). Additionally, the clinical presentation of our cat was unusual as it included decreased vmROM, signs of rhinitis/pharyngitis, multiple cranial nerve deficits, and lacked locomotory and other neuromuscular abnormalities. Nerve deficits were reported in one other cat,¹² and chronic neurological presentations have been documented in 51% of people with FOP.¹⁷ Progression of the lesions commonly follows a cranial-to-caudal and axial-to-appendicular direction in people, sparing the diaphragm, tongue, extra-ocular, smooth and cardiac muscles.² However, retrograde involvement, from hindlimbs to forelimbs, has been described in cats.^8,10,11,14

Previous feline cases of FOP-like disease have shown a poor prognosis owing to rapid progression of the disease, generally leading to a decision of euthanasia a few weeks or months after initial presentation.^{89101112131415–16} Cardiorespiratory failure, secondary to increased pulmonary vascular resistance, is the most common cause of death in human patients,¹⁸ usually occurring during the second to fifth decades of life.^1,2 Involvement of the oromaxillofacial region is an additional concern, as it may compromise vital primary requirements, such as air and food intake.

Differential diagnosis for restricted mouth opening in dogs and cats includes temporomandibular joint disease (fracture/luxation, dysplasia, osteoarthritis, neoplasia), any painful condition related to the masticatory region (eg, retrobulbar abscess, severe middle-ear disease) and systemic diseases such as tetanus.¹⁹ Masticatory myositis, a well-known autoimmune condition in dogs, was also recently described in a cat and is mainly characterised by its chronic presentation of bilateral symmetrical masticatory muscle atrophy.²⁰ Normal vmROM has been recently established in cats at a mean ± SD of 62 ± 8 mm (median 63 mm, range 41–84 mm);²¹ our cat had a vmROM of 22 mm when first presented.

In humans, the most important imaging feature of FOP is the late ectopic ossification of soft tissue that is easily seen with CT.^2,22,23 Non-mineralised early-to-intermediate - stage FOP lesions are oedematous, markedly inflammatory and progressively fibroproliferative. With CT, such lesions may appear as swelling of the facial plane around muscles that is isodense to hypodense with a variable enhancement pattern, depending on the balance between inflammation and neovascularity.

In our case, the hypodense areas within the masticatory muscles most likely represented areas of oedema and inflammation. The lack of enhancement after intravenous injection of iodinated contrast medium may indicate the absence of significant neovascularisation.

In the previous cases of feline FOP-like disease, CT was helpful for the characterisation, localisation and histopathological sampling of the soft tissue calcifications.^13,16 The degree of soft tissue calcification in our case, as assessed with CT, was considered low when compared with previously published CT images of cats. Considering the location of the lesions and their relative low degree of calcification it is likely that these lesions would have been missed on plain radiographs. Heterotopic new bone formation in human FOP and feline FOP-like disease involves various connective tissues, such as the muscles, tendons, ligaments, aponeuroses, fasciae and joint capsules.^1,2,16 In people, MRI and (fluorodeoxyglucose)-positron emission tomography are more sensitive than CT in the early diagnosis of FOP and in recognising ‘flare ups’.^2,22,23

Mandibular periostitis ossificans is a condition that has been described in young people and dogs, characterised by periosteal new bone formation, without any intramuscular heterotopic lesions, and mostly related to mandibular maturation.²⁴ Lesions are mainly distributed along the caudal ventral border of the body of the mandible, and are suspected to be associated with dental disease in young patients.²⁴

Periosteal bone proliferation in the present case follows the muscle attachments along the temporal and parietal bones and the ramus while sparing the caudal ventral border of the body of the mandible. Similar descriptions involving various locations, including the masticatory muscle attachments, are well described in humans with FOP.^2,25 Such lesions were not specifically reported in cats with FOP-like disease,^{89101112131415–16} but most likely correspond to early manifestations of the extra-skeletal bridging ossification process in our case.

Low-grade extra-skeletal osteo-/chondrosarcoma share some histological features with FOP that may lead to a challenging histopathological differentiation of these two conditions.²⁶ In our case, a malignant bone tumour could have been considered based on ectopic bone formation within skeletal muscle, cellular atypia and proliferation. However, the bilateral distribution, low mitotic index and concentric organisation with very - well-differentiated bone spicules led to us ruling out malignancy.

Periosteal new bone formation was described in human and canine myositis ossificans circumscripta.^{22232425262728–29} The lack of consistent inflammatory features on biopsied material and the bilateral multifocal asymmetric distribution of lesions are not compatible with myositis ossificans circumscripta.^22,2728–29

The exclusive involvement of masticatory muscles might have suggested a specific inflammatory response against type 2M myofibres. This entity was ruled out owing to the absence of characteristic histological features. Although masticatory muscle myositis might be associated with dystrophic mineralisation in necrotic myofibres, it is not associated with heterotopic new bone formation among myofibres.²⁰ However, the 2M antibody test may be useful, as a possible non-invasive diagnostic test to differentiate masticatory muscle myositis from FOP-like disease.

The cause of the multiple cranial nerve deficits in our case remains undetermined. As previously described in human FOP, and cats with FOP-like disease, some of these deficits may have been due to the entrapment of vessels and nerves by the lesions, lymphedema, demyelisation or venous thromboembolism.^2,9,16 Additionally, prolonged trismus and muscle swelling may have impaired the blood supply of the maxillary arteries to the retinae and brain.^3031–32

In people, FOP lesions may arise after a regional inflammatory process such as a viral infection, fatigue, or muscle trauma and strain.¹ No history of trauma was reported in our case. The chronic rhinitis/pharyngitis may have been an inflammatory trigger leading to metaplastic changes within the masticatory muscles. Experimental evidence in mice and primates suggests there is an upregulation of BMP-1 receptors secondary to respiratory epithelium inflammation.^33,34 Through their chemotactic activity, BMPs are not only responsible for bone deposition,³⁵ but could also play a key role in the respiratory inflammatory process.^33,34 Upper airway respiratory disease in our patient may have triggered or been a consequence of BMP overexpression.

Mutation of ACVR1/ALK2 has been shown to trigger BMP-1 receptor constitutive activation, resulting in osteoblast differentiation and bone formation in most human cases of FOP and in one cat with the disease.^14,35 However, a detailed genetic background, as well as the material and methods used for the genetic characterisation of the disease, have, to date, not been reported in cats. Using sequence alignment, a comparison between the well-documented Homo sapiens and Felis catus non-documented ACVR1 showed that they were quite different in their exonic structure and sequence. Thus, it was not possible to use H sapiens ACVR1 to design primers for F catus ACVR1 gene analysis. Sequencing the whole ACVR1 gene in our case permitted the characterisation of two deletions in intron 8, upstream of the last coding DNA sequence, constituting exon 9.

Probably the most interesting mutation corresponded to a four-base TTGA deletion, from a motif very similar to the ‘branching site necessary for intron splicing’ or ‘splice-site mutation’ that is (C/T)TNA(C/T) or YTNAY in humans.³⁶ The latter are located at the very termini of introns and drive exon recognition. Such genetic mutations occur in the very specific site at which splicing takes place during the processing of a precursor mRNA into a mature mRNA. The deletion of the splicing site may result in remaining mature mRNAs that may lead to the production of abnormal proteins. Genotype variations from classic FOP have also been described in humans.³⁷

Despite the fact that no direct link can be established between the mutation observed and FOP, it is possible that the splice site mutation of ACVR1 observed in our case could be responsible for the disease. Full confirmation would require functional analysis of the variant or, failing that, bioinformatic predictive algorithms.³⁶ This condition was considered very likely in regard to the clinical presentation, and complementary results, similar to some oromaxillofacial descriptions of FOP in humans.^3456–7

Conclusions

Although uncommon, heterotopic ossification-related diseases such as FOP should be included in the differential diagnosis of restricted-mouth-opening disorders in cats. Suspicion should be raised following clinical and diagnostic imaging findings. Any additional trauma, such as muscular biopsies or surgical procedures, should be minimised as they may activate bone deposition and worsen muscular lesions. Studies are needed in order to further characterise the genetic background of the disease, promote the development of a commercially available test and offer a non-invasive diagnostic tool.