Multiple Epiphyseal Dysplasia

Multiple epiphyseal dysplasia is one of the most common skeletal dysplasias and a leading genetic cause of early-onset osteoarthritis. It is frequently misdiagnosed as bilateral Perthes disease, and that misdiagnosis leads to inappropriate treatment. Recognising it early allows genetic counselling, avoids unnecessary surgery in childhood, and sets realistic expectations for the joint replacement that is likely in early adulthood.

Who. Prevalence is approximately 1 in 10,000 live births, worldwide and with no ethnic predilection. Inheritance is predominantly autosomal dominant, with rare recessive forms, and the dominant forms affect males and females equally. Penetrance is variable, and some carriers are asymptomatic.

What it costs. Short stature is mild, usually 2-3 SD below the mean, and not as severe as achondroplasia. The real burden is progressive pain and stiffness in the hips, knees and ankles, and joint replacement is often required at 30-50 years.

The matrix. Normal articular cartilage relies on the coordinated function of collagen II, the structural framework; collagen IX, which cross-links the fibrils and stabilises them; and COMP, a secreted calcium-binding protein that facilitates collagen II and IX fibrillogenesis. Mutations in COMP, COL9A1-3 or MATN3 disrupt the assembly of this extracellular matrix, and every MED gene acts here.

COMP. Mutant COMP misfolds and is retained in the rough endoplasmic reticulum of the chondrocyte, triggering an ER stress response and premature chondrocyte apoptosis. Pseudoachondroplasia is the same mechanism in the same gene, so the two are grouped as "COMPopathies"; pseudoachondroplasia is the more severe of the two, adding metaphyseal change and platyspondyly to the epiphyseal dysplasia seen here.

Collagen IX. The normal molecule is a heterotrimer of alpha-1, alpha-2 and alpha-3 chains that cross-links collagen II fibrils. Mutation disrupts heterotrimer assembly or fibril cross-linking, leaving a weak matrix with reduced tensile strength. COL9 disease tends to be milder than COMP disease but still causes early osteoarthritis.

The result. Endochondral ossification is defective: the epiphyses ossify late and irregularly, producing small, misshapen articular surfaces that are biomechanically inferior and predispose to early osteoarthritis.

View larger

By gene. Six numbered subtypes, one per gene. COMP is the commonest and tends to be the more severe, MATN3 the milder, and the recessive form is SLC26A2.

Genetic classification guides prognosis and genetic counselling but does not currently change orthopaedic management, which remains supportive.

Myopathy belongs to type IX collagen and C-terminal COMP, not to COL9A1. The association is commonly misstated. Mild myopathy has been reported with a COL9A3 mutation, with two families carrying C-terminal COMP mutations, and, in Jackson's 2010 series, with COL9A2 mutations in two families (Am J Med Genet A 2010;152A(4):863-9, PMID 20358595). A child with MED who has proximal weakness or a raised CK points to type IX collagen or a C-terminal COMP mutation.

Osteochondritis dissecans. Those same COL9A2 families had osteochondritis dissecans. It is an under-remembered association of type IX collagen MED and is orthopaedically actionable: a knee that locks or gives way in a child with MED deserves imaging for a loose body rather than being written off as part of the dysplasia.

The recessive and dominant split. The European Skeletal Dysplasia Network's 130-patient analysis is the reference point. The largest proportion of autosomal dominant MED is COMP, with MATN3 and type IX collagen accounting for the rest, while autosomal recessive MED appears to result exclusively from SLC26A2 (Jackson, Hum Mutat 2012;33(1):144-57, PMID 21922596). "Recessive MED" and "SLC26A2" are, on current evidence, the same statement.

By severity. Severity correlates broadly with genotype (COMP tends more severe, MATN3 milder) but is variable within families due to penetrance.

Fairbank and Ribbing. Long before genetic subtyping, MED was split clinically into two eponymous phenotypes that examiners still use:

• Fairbank type, the more severe, generalised form: many joints involved (hips, knees, ankles, shoulders and hands), more pronounced short stature, and small epiphyses throughout the skeleton
• Ribbing type, the milder form: predominantly the hips, near-normal stature, often presenting only in adolescence or adulthood as premature hip osteoarthritis

The split is descriptive and overlaps with genotype (COMP tends toward the severe Fairbank pattern; the milder COL9 and MATN3 forms toward Ribbing), but it remains a useful clinical shorthand and a frequent viva term. Both share the cardinal MED features, bilateral symmetric epiphyseal dysplasia with a spared spine.

History. The child presents between 2 and 10 years of age with a waddling gait or joint pain, and the triad to recognise is short stature, waddling gait and joint stiffness or pain. The pain is progressive, worse with activity and easing with rest; morning stiffness improves with movement; at sport the child struggles to keep up with peers and tires early. The family history is often positive, so ask about relatives with early arthritis or joint replacement.

Examination. The gait is waddling, antalgic or Trendelenburg, the last from hip involvement. Stature is mildly short and proportionate. By region:

• Hips: restricted internal rotation, flexion and abduction; pain on log roll
• Knees: varus or valgus deformity, restricted flexion, crepitus
• Ankles: stiff subtalar joint, possible equinus
• Hands: short, stubby fingers (brachydactyly) in some subtypes
• Spine: essentially normal, with no significant kyphosis or scoliosis

The hip-first trap. Many patients present with hip pain and bilateral femoral head abnormalities on the pelvis film, and the default registrar diagnosis is bilateral Perthes. The two are separated on the following grounds, and the distinction matters because Perthes is treated with containment (bracing, osteotomy) while MED requires supportive care and genetic counselling:

• Temporal pattern: in MED both femoral heads are small at the same time; in bilateral Perthes the second hip follows the first by 1-3 years and the two sides are at different stages
• Family: MED is familial (usually AD); Perthes is sporadic with no family history
• Acetabulum: normal coverage with no acetabular changes in MED; acetabular dysplasia and changes develop in Perthes
• Other joints: knees, ankles and shoulders are also affected in MED; Perthes is isolated to the hips and does not affect the knees or shoulders
• Onset: typically 2-10 years in MED, 4-8 years in Perthes
• Spine: normal in both; significant platyspondyly is neither, and should make you think of spondyloepiphyseal dysplasia

The critical test. Always look at the knee and ankle films before committing to Perthes. If the distal femoral and proximal tibial epiphyses are also small and irregular, the diagnosis is MED.

Radiographs. The skeletal survey is the cornerstone of diagnosis, and the pattern of epiphyseal involvement across multiple joints is what makes it. Request:

• AP pelvis
• AP and lateral of both knees
• AP and lateral ankles
• Lateral thoracolumbar spine
• PA chest
• AP and lateral of the wrists, for bone age

The features to look for are small, irregular, delayed epiphyses at multiple joints; bilateral and symmetric involvement; preserved acetabular depth and coverage; a normal or near-normal spine with normal vertebral body height; and delayed bone age. Some subtypes add short metacarpals and phalanges.

View larger

Genetic testing. Indicated once MED is confirmed or suspected on clinical and radiographic grounds, by targeted gene panel or whole exome sequencing covering COMP, COL9A1, COL9A2, COL9A3, MATN3 and SLC26A2. Approximately 50-60 percent of clinically diagnosed cases have an identifiable mutation in one of these genes.

MRI. Its role is surgical planning rather than diagnosis: assessing articular cartilage integrity in symptomatic joints and planning joint preservation surgery. It shows thinning or loss of articular cartilage, subchondral cysts, osteophytes and marrow oedema in early osteoarthritis.

Blood tests. TSH, free T4, calcium, phosphate, alkaline phosphatase and vitamin D are normal in MED. They are sent to exclude metabolic and endocrine causes of epiphyseal dysplasia (hypothyroidism, and as a skeletal dysplasia screen).

The double-layered patella. A horizontal lucent cleft dividing the patella into anterior and posterior ossified layers on the lateral knee radiograph is a rare but near-specific sign of recessive MED caused by SLC26A2 (DTDST) mutations (rMED, EDM4), reflecting the abnormal endochondral ossification of the patellar cartilage. Recessive MED also classically pairs the epiphyseal dysplasia with other DTDST-spectrum features: clubfoot, cleft palate, brachydactyly, and ear swelling or deformity. In a short-statured child with epiphyseal changes, this sign points specifically to the recessive form rather than the dominant COMP and COL9 types.

Childhood (age 2-16). The goals are to maintain mobility, manage pain, monitor epiphyseal development and provide genetic counselling. The mainstay is non-operative.

At diagnosis. The child is taken on by a multidisciplinary team of paediatric orthopaedics, genetics and physiotherapy. Genetic counselling covers the inheritance pattern, variable penetrance and the implications for family planning. Activity is modified away from high-impact sport towards swimming and cycling, and physiotherapy maintains range of motion, strengthens the periarticular muscles and trains the gait.

Annual review. Height, gait, joint range of motion and pain are recorded, with annual AP pelvis and knee radiographs to track epiphyseal development. Shoe modifications or ankle-foot orthoses address ankle involvement or leg-length discrepancy. Pain is managed with activity modification, NSAIDs as needed and weight optimisation.

Surgery in childhood. Reserved for significant angular deformity or mechanical axis deviation:

• Corrective femoral or tibial osteotomy for varus or valgus greater than 15 degrees, to improve joint biomechanics
• Hip arthroscopy has a limited role and may be considered for labral tears or impingement in adolescents
• Epiphysiodesis is rarely indicated, because the leg-length discrepancy in MED is usually mild

Adulthood (age 16 and older). The goals shift to managing early-onset osteoarthritis, timing joint replacement well and maximising quality of life.

Early adulthood (16-30). Conservative care continues: weight management, low-impact exercise (swimming, cycling, elliptical training, avoiding running and contact sports) and physiotherapy, with paracetamol, NSAIDs and intra-articular corticosteroid injections for flares. Serial radiographs track the progression of osteoarthritis.

Joint preservation (20-40). Arthroscopic debridement treats mechanical symptoms from loose bodies or labral tears. A valgus-producing proximal femoral or high tibial realignment osteotomy suits carefully selected patients with preserved joint space and a correctable deformity. PRP and mesenchymal stem cell injections have emerging evidence but no proven disease-modifying effect in MED.

Arthroplasty (typically 30-50). Total hip arthroplasty is the most common operation; outcomes are generally good, but stem and cup sizing can be challenging because the femoral and acetabular anatomy is altered. Total knee arthroplasty may be needed for end-stage knee osteoarthritis, and standard implants are usually appropriate. Counselling covers implant longevity (15-20 years for conventional bearings), activity restriction after arthroplasty, and the likelihood of revision within the patient's lifetime.

What the MED-specific data add. The generic warning about young-age revision should be read against the largest series of THA in MED (Park 2025, 88 hips followed a mean of 16 years): survivorship free of any revision was 95.9% at 16 years, with the Harris Hip Score rising from 43 to 91, no dislocations, and an 8% intraoperative periprosthetic femoral fracture rate from the narrow femur and reduced bone quality. Its mean age at surgery was 50, so it does not directly describe a patient replaced in their early thirties, but it is the durability evidence to quote rather than generic young-arthroplasty figures.

The primary morbidity is early osteoarthritis. Unlike Perthes, where the goal is to preserve the shape of the femoral head through containment, in MED the cartilage matrix is inherently defective and no treatment prevents the progression to osteoarthritis. The surgeon's role is to optimise joint biomechanics by correcting deformity, to get the most out of non-operative management, and to time joint replacement well, balancing the patient's pain and disability against the finite lifespan of a prosthesis in a young, active patient.

By genotype. Onset of osteoarthritis and age at first arthroplasty track the gene.

What shifts the prognosis. The best outcomes go with MATN3 and COL9 mutations, milder epiphyseal irregularity, compliance with activity modification and a healthy BMI. The poorer outcomes go with COMP mutations and severe epiphyseal irregularity, a high BMI (which accelerates osteoarthritis), high-impact activity, and onset of symptoms before age 5.

What to tell the patient. MED does not affect life expectancy, intelligence or systemic health; the morbidity is confined to the musculoskeletal system. With modern joint replacement, most patients achieve good function and independent living.