Cozen Fracture

The phenomenon. Leonard Cozen described it in 1953: a proximal tibial metaphyseal fracture in a young child, usually unremarkable in itself, followed months later by a progressive valgus deformity of the tibia. The fracture is metaphyseal, not physeal, which is why Salter-Harris grading does not apply to it, and the valgus appears after the fracture has united, often despite a perfectly healed fracture. Distinguish it from Blount disease, from physiological genu valgum and from a true physeal bar after a proximal tibial physeal injury; the discriminators are tabulated under Investigations.

Why it matters. The Cozen phenomenon is one of the most important and frequently missed concepts in paediatric trauma, because a seemingly trivial fracture can, months later, produce a striking deformity that alarms both the parents and the treating surgeon. Understanding that the deformity improves substantially with growth is the difference between appropriate reassurance and inappropriate early surgery; recurrence after an early osteotomy is one of the most-cited pitfalls in paediatric orthopaedic textbooks.

A naming trap worth ten seconds. There are two unrelated Cozen eponyms in orthopaedics and they share nothing but a surname. This one is Leonard Cozen's 1953 description of post-traumatic tibia valga after a proximal tibial metaphyseal fracture in a child. Cozen's test is a provocative test for lateral epicondylitis - resisted wrist extension reproducing lateral elbow pain, a different Cozen, a different limb and a different decade of adult life. If a question says "Cozen", read the rest of the stem before answering.

Who. Children aged 3-8 years at peak, with most series spanning 2-10 years and some reporting cases up to age 10. There is a slight male predominance (1.2-1.5:1) that reflects exposure to play and sport; either side is affected without strong predominance, and bilateral cases are exceedingly rare. The proximal tibial physis is highly active in this age group, and that activity contributes to the asymmetric growth that follows.

How common. Proximal tibial metaphyseal fractures are among the most common paediatric lower-limb fractures and are reported worldwide; the mechanism varies by environment, with traffic injuries in low- and middle-income countries and sport in high-income ones. A substantial proportion of these children go on to valgus, but the exact figure is not established, and the numbers that circulate are examined under Outcomes and Prognosis.

Mechanism of injury. The classic mechanism, described by Cozen, is a valgus force on an extended knee. The others are a fall from a height onto a planted foot with the knee extended, a twisting or abduction force from a bicycle or scooter, a valgus blow in football, skiing or a playground fall, and a dashboard-type injury to an older child travelling as a passenger. The mechanism predicts the pattern, and the pattern predicts the late valgus risk, so document it clearly: it is part of the medicolegal record.

View larger

No single mechanism explains it. More than one is likely operative in any given child, and the most-cited theories are four. What they share is the implication that the growth disturbance may persist for years, which is the critical reason to defer a corrective osteotomy.

Asymmetric growth stimulation (Cozen, 1953). The hyperaemia of fracture healing stimulates the medial proximal tibial physis and produces relative medial overgrowth. The support is Cozen's original clinical series and the consistent timing of the valgus after union, and the cleanest mechanistic evidence is Green's 1983 case report, in which growth arrest lines in the proximal tibial metaphysis showed that the valgus after a nondisplaced fracture healed in normal position came from overgrowth of the medial physis, not from lateral arrest or malunion.

Tibial tubercle apophysis overgrowth (Ogden). The tibial tubercle is a continuation of the proximal tibial physis, and overgrowth of the lateral tibial tubercle apophysis contributes a valgus moment generated distal to the physis. The support is anatomic and radiographic study of the apophysis.

Soft-tissue interposition (Skak, 1982). Interposed periosteum or pes anserinus tendon on the medial side, with an intact fibula acting as a strut. Skak's seven children had primary valgus angulation with medial cortical diastasis, six with radiological signs of soft tissue interposed medially in the fracture gap, and he concluded that routine surgery to remove the interposed tissue is not justified. Lateral tethering by the fibula or the pes anserinus is a different proposed mechanism and should not be cited to his paper.

Occult crush of the lateral physis (Salter-Harris V). A sub-radiographic crush injury to the lateral physis slows lateral growth, so the valgus emerges with growth from an injury that was undetected at presentation. MRI studies show lateral physeal signal change in some cases.

Why the deformity is often transient. The valgus usually appears 6-12 months after injury, peaks around 9-18 months and then remodels, most within 2-3 years and without intervention. As long as the proximal physis is open and the child has significant growth remaining, the asymmetric growth can re-equilibrate; skeletal immaturity is the best friend of the patient with Cozen valgus.

Why early osteotomy fails. The growth disturbance that produced the valgus persists for years, so an osteotomy performed while it is still active invites recurrence. A rate of up to 50 percent is conventionally quoted, though that figure is not established in the primary literature; what the largest published series records is that corrective osteotomy has been abandoned because of frequent complications including recurrent valgus. A child with a recurrence may end up with both an osteotomy scar and a worse deformity than the original, and each osteotomy adds scar, infection risk and stiffness to a problem that nature was correcting.

The acute fracture. Four patterns, and the late valgus risk runs paradoxically the wrong way: it is higher in the greenstick and less-displaced variants than in a displaced fracture that is anatomically reduced, and valgus can follow even an undisplaced injury. This is a recurring exam point.

The evolving deformity. The clinical course is a dynamic remodelling process, not a one-off surgical decision, and it is staged by time from injury.

History. Establish the mechanism, whether a valgus force, an axial load or a twist, and the energy: a low-energy playground fall or a high-energy fall or vehicle collision. Inability to bear weight is usually present acutely, and knee pain and swelling may dominate the presenting complaint. Always document pre-injury alignment if it is known, and any prior injury or deformity.

Examination. Inspect for bruising, swelling, angular deformity and an open wound; the tenderness is localised to the metaphysis, not the joint line. Document distal pulses, capillary refill, sensation and motor function, the peroneal nerve especially, and review the compartments serially for the first 24-48 hours, particularly after casting. Knee ligaments are difficult to assess acutely; examine under anaesthesia if a reduction is required.

Measuring the alignment. Four measurements track the deformity, from the goniometer in clinic to the long-leg film, and the mechanical axis is the most reliable of them.

Compare with the other side. Children have a wide range of physiological valgus, up to 7 degrees between ages 3 and 6, so the most useful finding is asymmetry between the injured and uninjured sides. A 10 degree valgus on the injured side with a 5 degree valgus on the other is more significant than a 7 degree valgus bilaterally.

Radiographs. The diagnosis of the Cozen phenomenon is clinical-radiographic, not MRI-dependent, and plain films are the cornerstone. Acutely, AP and lateral views of the tibia including the knee and ankle confirm the fracture and show displacement, angulation and fibular involvement, with an AP pelvis-to-ankle long-leg film added in older children for deformity assessment. At follow-up the films are for measurement: tibio-femoral angle, mechanical axis, MPTA and LDFA.

At union. Radiographs at cast removal, repeated at 6 weeks if pain persists, confirm union, document any early deformity and provide the baseline for the 6-month review. Look for early tibial asymmetry, physeal widening and tubercle overgrowth.

When valgus develops. A standing long-leg film is the gold standard for measuring the mechanical axis and tracking the deformity in the growing child: it documents the deformity quantitatively and gives the mechanical axis deviation. Repeat it every 6 months until the deformity resolves or reaches a clear plateau.

MRI of the proximal tibial physis. Reserved for the deformity that is severe or progressive, or that fails to remodel, to look for an occult lateral physeal bar or a Salter-Harris V injury. The sequences are coronal T1 and gradient echo or PD fat-saturated through the physis, and young children often need sedation.

CT with 3D reconstruction. For pre-operative planning of complex or recurrent deformity. It defines the bony anatomy for osteotomy design, whether closing wedge, opening wedge or dome.

The differential. A tibial deformity after a fracture has a short list of alternatives, and the discriminators are the history of a preceding fracture and the symmetry of the finding.

The acute fracture. The goal is union in acceptable alignment and a family who understand the natural history.

Reducing the greenstick in valgus. Many paediatric orthopaedic surgeons complete the fracture, reversing the intact cortex, and reduce the tibia in slight varus. This is thought to lower the rate of post-traumatic valgus by allowing more uniform physeal stimulation, and slight over-correction may reduce the late valgus. Document the reduction manoeuvre clearly.

Counselling at the first visit. This is the most important single conversation of the episode. Tell the family that a knock-knee appearance may develop on the injured side months after the fracture has healed, even after perfect healing; that it is likely to self-correct with growth; that the appropriate response is observation and review, not surgery; and that follow-up continues until remodelling is complete.

Document it. Write a brief clinic note that the family was counselled about the Cozen phenomenon, the natural history and the rationale for observation. The note is medicolegally essential, it pre-empts complaints, and it protects the surgeon, the family and the child; skip the conversation and you will be the surgeon asked to "fix" a valgus that was always going to correct, and may end up doing an osteotomy that then recurs.

When valgus appears. From 6-12 months the goal is to avoid premature surgery: document, reassure and monitor. A deformity that is getting worse at the 6-month review is the natural course rather than a failure - in Skak's series the valgus progressed through the first year despite clinical healing and slowly regressed from one to two years after injury - so the default for an immature child with emerging Cozen valgus is observation, and surgery is the exception, not the rule.

Surgical options. Surgery is reserved for persistent deformity: valgus that persists or progresses beyond skeletal maturity, severe functional impairment, or pain. In a still-growing child the trigger is persistent, symptomatic valgus that is not remodelling, by the thresholds in the table above, and the operation is chosen by the growth remaining: guided growth while the physes are open, osteotomy at or near maturity.

Guided growth (hemi-epiphysiodesis). A tension-band plate (8-Plate) or staple across the medial proximal tibial physis slows medial growth and lets the lateral side catch up. It is for the skeletally immature child with persistent valgus and at least 1-2 years of growth remaining, estimated from skeletal age and remaining-growth charts or the multiplier; with too little growth left it will not correct in time, and osteotomy at maturity is the answer. The attractions are a small incision, a day-case procedure, no osteotomy, and reversibility, the plate being removed after correction; the costs are the need for residual growth, implant-related irritation and rebound.

Why it beats osteotomy in the growing child. It works with the remaining growth rather than against a persisting growth asymmetry, so it does not carry the recurrence problem of osteotomy before maturity. It is not free of recurrence, though: in the largest series, correction of the mechanical axis and MPTA was achieved in all but one of 19 patients, there were five recurrences in four patients, four of which corrected with repeat tethering, and two patients overcorrected because the family did not return for timely follow-up.

Rebound and the timing of removal. Once the plate is removed with the limb straight, the previously restrained physis can resume asymmetric growth and the valgus can recur; removing the plate too early is the classic cause of recurrence. Monitor after removal, and in a young child consider slight overcorrection before removal or simply leave the implant until closer to maturity. The tether must come out in the end, though, or the leg overcorrects, which is why the operation needs a family who will attend follow-up.

Corrective tibial osteotomy. A closing-wedge or opening-wedge osteotomy of the proximal tibia, sometimes with a fibular osteotomy, realigns the mechanical axis. It is for the skeletally mature patient, or one approaching maturity with severe deformity, and should only be performed when the growth plates are nearly closed or the family accepts the risk of recurrence. The risks are recurrence, especially if performed too early, compartment syndrome, neurovascular injury, non-union and hardware irritation. Measure limb length at the same time, for the reason given under Outcomes and Prognosis.

Combined and salvage procedures. For recurrent valgus after osteotomy, multiplanar deformity or an associated limb-length discrepancy, the options are multi-level osteotomies, gradual correction with an external fixator (Taylor Spatial Frame), or hemi-epiphysiodesis followed by osteotomy in staged fashion. Each additional procedure adds scarring, stiffness and infection risk; avoid accumulating operations on a child.

The valgus deformity heads the list; its prevention is the counselling and surveillance described under Management.