Epiphysiodesis for Leg-Length Discrepancy

Indication. A symptomatic or progressive leg-length discrepancy predicted at skeletal maturity of roughly 2 to 5 cm, in a skeletally immature child with enough growth remaining in the long limb to make up the difference, where the family accepts the loss of final height in exchange for avoiding a lengthening. The short limb is left alone — epiphysiodesis SHORTENS the long limb and lets the short one catch up. The ideal candidate has a predicted discrepancy in the 2 to 5 cm window, a bone age before the end of growth (broadly girls up to about 14 years and boys up to about 16 years bone age, but always calculated individually), and serial measurements that project reliably to maturity. Relative indications. A discrepancy of less than 2 cm rarely needs surgery — a shoe raise or observation is usually sufficient. A 2 to 2.5 cm discrepancy may be managed non-operatively; epiphysiodesis is offered if it is symptomatic or clearly progressive. A mild discrepancy in a child with an underlying syndrome, where lengthening morbidity is undesirable, is another relative indication. Contraindications. Absolute: skeletal maturity already reached (the physis is closed, so there is no growth left to arrest); a predicted discrepancy greater than 5 cm where shortening the long limb would leave the patient unacceptably short (favour lengthening); and active infection at the operative site. Relative: very large discrepancies better served by lengthening (or combined lengthening plus contralateral epiphysiodesis); significant short stature where any height loss is poorly tolerated; and uncertain or unreliable growth prediction from insufficient serial data or an ambiguous bone age. Why timing is everything. The whole operation succeeds or fails on doing it at the moment when the growth the long limb has left to give exactly equals the discrepancy to correct. Arrest is permanent and immediate — the physis stops growing from the day of surgery. Too early and the long limb stops while too much growth remains, so the short limb overtakes it (overcorrection); too late and too little growth remains to make up the difference (undercorrection). A one-year error in bone age alters the final correction by roughly 1 cm at the distal femur.

Consent specifically for the irreversible nature of the arrest, the risk of over- or under-correction from mistiming, angular deformity from incomplete or asymmetric arrest, the loss of final height (the child will be shorter than their genetic potential and the operated knee will sit at a slightly lower level), infection, and — for a proximal tibial arrest — common peroneal nerve injury. Setup. Supine on a radiolucent table with the affected (long) limb prepared free and the image intensifier brought in from the opposite side (a small bump under the ipsilateral hip helps for lateral imaging). General anaesthesia; a thigh tourniquet may be applied but is not always required for percutaneous work. Image intensifier in BOTH AP and lateral is mandatory, because the physis is not flat.

The goal is to ablate the targeted physis completely and symmetrically so that bony bridging occurs and the long limb stops growing, while protecting the joint surface, the neurovascular structures, and (at the proximal tibia) the common peroneal nerve. The dominant modern technique is the percutaneous drill-and-curette epiphysiodesis (Canale and Bowen) — minimally invasive, low morbidity, day-case — laid out step by step below; the open Phemister and percutaneous transphyseal screws (PETS) are the alternatives.

View larger

View larger

Open Phemister (classic). The original described technique — reliable but more invasive than percutaneous methods. Through medial and lateral longitudinal incisions over the physis, expose the plate and cut a rectangular block of bone spanning from epiphysis to metaphysis with an osteotome, curette the remaining physeal cartilage from the bed, then rotate the bone block 90 to 180 degrees and replace it so cortical bone bridges the physis and promotes bony arrest, and close in layers. Phemister gives direct visual confirmation of physeal removal but at the cost of larger incisions and longer recovery than the percutaneous drill-and-curette method. Percutaneous transphyseal screws (PETS). Cannulated screws are placed obliquely across the physis — typically two screws, medial and lateral — to bridge and arrest the growth plate. The screws are inserted percutaneously under fluoroscopy from the metaphysis, across the physis, into the epiphysis; the attraction is a technique that is theoretically more controllable and described as relatively reversible if the screws are removed early before the bridge consolidates, though true reversibility is debated, and the hardware can later be removed. The key outcome caveat is that arrest with PETS is GRADUAL rather than immediate and is frequently incomplete: in the Ilharreborde maturity series the final loss of growth was only about 66 percent of that predicted pre-operatively, so the planned correction should be OVER-estimated to compensate. Crucially, tibial PETS produced a valgus deformity in 20 percent of patients and a high revision rate, leading those authors to ABANDON PETS in the proximal tibia while continuing to use it in the femur — in contrast to the immediate, complete arrest of a well-performed drill-and-curette epiphysiodesis.

View larger

Rehabilitation | Phase | Timing | Weight-bearing and wounds | Monitoring | |-------|--------|---------------------------|------------| | Early recovery | 0 to 2 weeks | Weight-bearing as comfort allows (percutaneous); protected weight-bearing if open Phemister. Simple dressings over stab incisions; absorbable sutures; simple analgesia. Return to school in 1 to 2 weeks for the percutaneous technique | Wound check | | Surveillance | Every 3 to 6 months until maturity | Normal activity | Standing long-leg radiograph or scanogram — confirm lengths are converging, watch for angular deformity, recalculate actual versus predicted | Expected outcomes. When timing and growth prediction are accurate, epiphysiodesis reliably equalises discrepancies in the 2 to 5 cm range with low morbidity and an excellent day-case recovery profile. The percutaneous drill-and-curette technique gives results comparable to the open Phemister technique with smaller scars and faster recovery, and is the most commonly used method today. The principal determinant of success is the accuracy of the timing decision, not the technique used — accurate timing with complete arrest gives reliable equalisation, whereas an arrest performed too early overcorrects (the opposite discrepancy), too late leaves a residual discrepancy, and an incomplete or asymmetric arrest produces angular deformity.

Aetiology of leg-length discrepancy. Common causes to identify during the work-up include congenital limb hypoplasia or hemihypertrophy, post-traumatic physeal arrest, previous infection involving a growth plate, neuromuscular conditions, skeletal tumours or their treatment, and constitutional or vascular anomalies. The aetiology determines whether the discrepancy is progressive and proportionate, which in turn governs which prediction method applies. Bone age assessment. Skeletal maturity — not the calendar — determines remaining growth. The Greulich-Pyle atlas compares a left hand and wrist radiograph to standard plates and is the standard method across childhood. The Sauvegrain (elbow) method uses ossification of the elbow during the pubertal growth spurt and is more discriminating around puberty, when timing decisions are most sensitive. A discrepancy of one to two years between bone age and chronological age is common and clinically important. Normal physeal growth rates. To remove length you arrest the physis (or physes) contributing the relevant growth. For a femoral-side discrepancy, arrest the distal femur; for a combined correction, arrest the distal femur AND proximal tibia on the long side. | Physis | Growth per year | Rule of thumb | Share of limb growth | |--------|-----------------|---------------|----------------------| | Distal femur | 9 to 10 mm | three-eighths inch | About 70 percent of femoral length; the fastest-growing physis in the body | | Proximal tibia | 6 mm | one-quarter inch | About 55 to 60 percent of tibial length | | Proximal femur | About 3 mm | — | Minor contribution | | Distal tibia | About 3 to 5 mm | — | Minor contribution |

View larger

Growth-prediction methods. Three validated methods are used; all require bone age, not chronological age, and a delayed or advanced skeletal age of one to two years substantially changes the remaining-growth estimate. - Green-Anderson growth-remaining charts are built on longitudinal data (Menschik) of normal limb growth. They plot the remaining growth of the distal femur and proximal tibia against skeletal age, so the surgeon reads off how much growth is left at a given bone age and chooses when to arrest to remove the desired length. Their strength is that they directly give growth-remaining figures; their weakness is that they assume the child follows a normal percentile.

• The Moseley straight-line graph transforms the non-linear growth data into a straight line so that serial measurements of the short and long limbs plot as lines, with skeletal-age reference lines drawn and the effect of epiphysiodesis shown as a change in slope of the long-limb line. It visually integrates multiple serial measurements and lets you simulate the timing of surgery, but needs at least two or three data points for accuracy.
• The Paley multiplier method uses a multiplier (derived from databases of normal growth, specific for age and sex) applied to the CURRENT discrepancy or current bone length to predict the value at maturity: the predicted discrepancy at maturity equals the current discrepancy multiplied by the multiplier (for congenital, proportionate discrepancies). It is a single time-point calculation with no need for serial data — quick and widely used — but assumes the discrepancy grows proportionately, so it is best for congenital causes.

Relevant surgical anatomy. The distal femoral physis is undulating, with anterior and posterior portions — the most complex physis, which is why open Phemister or percutaneous drilling must address its full three-dimensional extent; it is approached medially and laterally just above the joint line, anterior to the collateral ligament origins, and the peripheral perichondrial ring (ring of LaCroix and zone of Ranvier) must be ablated for complete arrest, since residual peripheral physis causes deformity. The proximal tibial physis lies just distal to the joint line and the tibial tubercle apophysis anteriorly is part of this growth complex; the common peroneal nerve wraps around the fibular neck laterally and is at risk with lateral proximal tibial dissection, and the proximal tibiofibular relationship matters because combined tibial epiphysiodesis often includes the proximal fibular physis to avoid a relatively long fibula. On fluoroscopy the physis is seen as a lucent transverse line in both AP and lateral views (it is not flat), and the curette and drill must reach the central, anterior and posterior portions, not just the easily accessed peripheral margins.

Additional references: - Métaizeau JP, Wong-Chung J, Bertrand H, Pasquier P (1998). Percutaneous epiphysiodesis using transphyseal screws (PETS). J Pediatr Orthop 18:363-9. PMID 9600565. — Original description of PETS; mean bone-length inequality fell from 2.47 cm to 0.51 cm at maturity across 32 cases.

• Bowen JR, Johnson WJ (1984). Percutaneous epiphysiodesis. Clin Orthop Relat Res 190:170-3. PMID 6488627. — Original clinical description of percutaneous drill-and-curette epiphysiodesis under image intensification.
• Phemister DB (1933). Operative arrestment of longitudinal growth of bones in the treatment of deformities. J Bone Joint Surg Am 15:1-15. — Original description of the open bone-block (Phemister) epiphysiodesis technique (predates PubMed indexing).
• Aguilar JA, Paley D, Paley J, et al. (2005). Clinical validation of the multiplier method, part II. J Pediatr Orthop 25:192-6. PMID 15718900. DOI 10.1097/01.bpo.0000150808.90052.7c. — Validation cohort: multiplier method mean error for residual discrepancy about 0.9 to 1.0 cm, more accurate than Moseley for residual LLD after epiphysiodesis.