Proximal Tibia Metaphyseal Fractures — Pediatrics

Proximal tibial metaphyseal fractures in children are deceptively problematic injuries. Although they appear minimally displaced or undisplaced on initial radiographs and often seem straightforward to manage, they are notorious for developing a progressive valgus deformity weeks to months after fracture — the Cozen phenomenon. Understanding the aetiology of this valgus overgrowth, the controversies surrounding prevention, and the long-term management of established deformity is essential for any paediatric orthopaedic surgeon.

• Demographics: most common in children aged 2–8 years; peak incidence 3–6 years; the greenstick or incomplete nature of the fracture in young children is the typical pattern; direct trauma (playground falls, bicycle-related) or valgus stress mechanism; the proximal tibial metaphysis in young children is a rapidly growing region with abundant vascularity — factors that contribute to asymmetric overgrowth
• Classification: proximal tibial metaphyseal fractures are typically described by their pattern — complete transverse, incomplete (greenstick — most common in this region), buckle/torus, or Salter-Harris physeal fractures (separate entity — epiphyseal fractures of the proximal tibia); the greenstick fracture pattern with an intact medial periosteum and cortex is the most commonly associated with subsequent valgus deformity; associated fibular fractures are less common than in diaphyseal fractures and their presence or absence may influence stability
• Initial displacement: paradoxically, initially undisplaced or minimally displaced fractures carry the highest risk of progressive valgus deformity — the `Cozen phenomenon`; moderate-to-severely displaced fractures that are reduced do not consistently develop the same post-fracture valgus overgrowth

• Definition: the development of progressive tibia valga (valgus deformity of the proximal tibia) following a proximal tibial metaphyseal fracture in a child, beginning weeks after injury and progressing over months to years; first described by Cozen in 1953; occurs even when initial reduction was anatomical or the fracture was minimally displaced; the deformity typically becomes apparent 3–6 months post-fracture and may continue to progress for 1–3 years before reaching a plateau or partially remodelling
• Magnitude: the valgus deformity can range from mild (5–10°) to severe (>30°); most commonly 15–25° of valgus develops; this represents a significant cosmetic and functional concern; the deformity is rarely associated with leg length discrepancy (both tibiae typically grow at similar rates) — this distinguishes the Cozen phenomenon from growth arrest, which causes both deformity and shortening
• Aetiology — the controversy: the mechanism of the Cozen phenomenon remains incompletely understood and several theories exist: (1) Periosteal tethering hypothesis — the intact medial periosteum at the fracture site tethers the medial cortex, reducing growth on the medial side while the lateral cortex grows unconstrained; (2) Interposition of soft tissue — the pes anserinus tendon insertion (sartorius, gracilis, semitendinosus) medially is disrupted or compressed by the fracture callus, creating asymmetric medial tethering; (3) Asymmetric stimulation of the proximal tibial physis — fracture haemarthrosis and hyperaemia selectively stimulate the lateral portion of the proximal tibial physis (closer to the fracture callus) causing accelerated lateral growth; (4) Inadequate reduction of medial comminution — any medial cortical impaction that is not recognised creates a medial height deficit, allowing relative lateral growth; the periosteal tethering and physeal stimulation theories have the most supporting evidence
• Natural history: approximately 70% of deformities spontaneously correct or improve significantly over 3–5 years through remodelling; children under 8 years with >3 years of growth remaining have the best remodelling potential; persistent deformity beyond age 8–10 years or deformity >20–25° is less likely to remodel and may require surgical correction

• The `slight varus` casting controversy: some surgeons deliberately reduce to slight varus or cast in slight varus to counteract the expected valgus tendency; others argue this is ineffective as the Cozen phenomenon occurs regardless of casting position and is driven by asymmetric physeal growth; no high-level evidence establishes that casting in varus prevents the Cozen phenomenon; however, it is widely recommended as a precautionary measure and logical given the expected deformity — avoiding any valgus in the cast is universally recommended; avoid any residual valgus in the fracture position as this will be additive to the subsequent Cozen valgus
• Surgical indications for the acute fracture: open fractures; irreducible fractures; neurovascular compromise; polytrauma (facilitating nursing and physiotherapy); associated ipsilateral femoral fracture (floating knee)

• Standing AP long-leg radiograph (hip-to-ankle): essential for measuring the mechanical tibiofemoral angle, the tibial deformity angle (TDA — the angle between the tibial axis and the joint line of the proximal tibia), and assessing whether deformity is in the tibia, femur, or both; the centre of rotation of angulation (CORA) — the level of maximum angular deformity — must be identified for accurate osteotomy planning; in Cozen phenomenon, the deformity is typically at the proximal tibial metaphysis itself
• Differential diagnosis of genu valgum in children: physiological valgus (peaks at age 3–4 years, resolves by age 6–7 years — bilateral and symmetric); Blount`s disease (tibia vara — opposite deformity); rickets (nutritional or renal — metabolic bone disease; widened physis on X-ray); enchondromatosis; skeletal dysplasias; post-fracture Cozen phenomenon (asymmetric, history of fracture); infantile tibia vara; distal femoral deformity

• Cozen phenomenon: progressive valgus deformity after proximal tibial metaphyseal fracture in children; begins 3–6 months post-fracture; most commonly greenstick/incomplete fracture; age 2–8 years; NOT caused by inadequate initial reduction — can occur after anatomical fracture
• Aetiology theories: periosteal medial tethering (most supported); asymmetric physeal stimulation by fracture haemarthrosis/hyperaemia; pes anserinus interposition; residual medial cortical impaction; NO single mechanism universally accepted
• Natural history: 70% correct spontaneously with growth; <8 years with >3 years growth = best remodelling; >20–25° or age >8–10 = surgical correction; observe on serial standing long-leg films every 6–12 months
• Acute management: long leg cast in slight varus (or at least neutral — avoid any valgus in cast); avoid residual valgus position (additive to Cozen valgus); 6–8 weeks immobilisation
• Guided growth (8-plate hemiepiphysiodesis): for progressive deformity >10–15° with >2 years growth remaining; tethers medial proximal tibial physis; gradual correction ~1–2° per month; reversible; most modern approach for growing child
• Corrective osteotomy: near or at skeletal maturity; >20–25° residual valgus; medial closing or lateral opening wedge; risk of post-osteotomy compartment syndrome
• Distinguish from physiological valgus: Cozen is asymmetric (one side), follows fracture, begins at 3–6 months; physiological valgus is bilateral, symmetric, peaks at 3–4 years, self-corrects by 6–7 years
• Standing long-leg AP film: essential; measure mechanical tibiofemoral angle, tibial deformity angle, CORA; assesses femoral vs tibial contribution to total deformity; plan osteotomy level based on CORA