We retrospectively reviewed fifteen U-osteotomy procedures performed for the correction of complex foot deformities. The most common deformities of the foot relative to the tibia were angulation in the frontal plane, translation in the sagittal plane, and limb-length discrepancy. Causes of these deformities included fibular hemimelia, diastrophic dysplasia, pilon fracture, arthrogryposis with postoperative deformity, rheumatoid arthritis, tibial hemimelia, subtalar joint coalition, malunited fractures, and clubfoot. During a twenty-one-year period (1988 to 2008), ten female and five male patients underwent U-osteotomy.
The U-osteotomy is performed through a modified lateral Ollier-type incision. It crosses the talar neck, passes through the sinus tarsi, and crosses under the sustentaculum tali to exit dorsal and posterior in the calcaneus (anterior to the Achilles tendon insertion). Radiographs were made and clinical examinations were performed on all fifteen patients preoperatively and at the time of final follow-up. The medical records of each patient were reviewed. Data regarding the type of deformity, etiology, patient age, concomitant surgery, duration of external fixation, duration of follow-up, previous surgery and other treatments, time to return to wearing normal footwear, and complications were extracted from the medical records.
The calcaneotibial angle was measured on a standing axial hindfoot alignment view radiograph as the angle between lines through the longitudinal axes of the tibia and the calcaneus. Normally, the longitudinal axis of the calcaneus should be in 0° to 5° of valgus relative to the longitudinal axis of the tibia. This radiograph was made with the patient standing on a radiolucent platform with the cassette angled 20°, ensuring inclusion of the entire calcaneus and the distal aspect of the tibia13.
The radiographic measurements obtained from the lateral radiograph of the foot included the calcaneal inclination angle, posterior translation of the calcaneus, and foot height14-17.The calcaneal inclination angle was calculated by drawing a line along the plantar cortex of the calcaneus and measuring its intersection with the weight-bearing surface (the floor). Posterior translation of the calcaneus was measured as the perpendicular distance between the line bisecting the longitudinal axis of the tibia and a point on the lateral process of the talus. The foot height was defined as the distance from the standing surface to the tip of the medial malleolus. Because the U-osteotomy is performed through the neck of the talus, we were unable to accurately measure the talus-first metatarsal angle on the anteroposterior or lateral radiographs.
Clinical Evaluation
Treatment outcome was evaluated with use of two standardized questionnaires: the American Orthopaedic Foot & Ankle Society (AOFAS) questionnaire18-20 and the Short Form-12 (SF-12) health survey21. The AOFAS questionnaire is a clinical outcome rating system designed to assess foot and ankle function. The maximum score is 100, with a higher score indicating a more beneficial postoperative outcome. This instrument is designed to assess pain (40 points), function including movement (50 points), and alignment of the hindfoot and ankle (10 points). The SF-12 health survey has physical and mental component summary scores that range from 0 to 100, with 0 indicating the lowest level of health and 100 indicating the highest level of health; the score is standardized, with a mean of 50 and a standard deviation of 10, on the basis of the 1998 general United States population.
Preoperative Planning
To accurately determine whether a patient is a good candidate for the U-osteotomy, appropriate radiographs must be obtained (Fig. 1)16. Since the U-osteotomy restores foot height and repositions the entire foot relative to the tibia, preoperative and postoperative full-length anteroposterior standing radiographs (including the pelvis, femur, tibia, and foot) need to be obtained to ensure equal limb length.
The most critical radiographs are lateral views of the foot and ankle. The lateral radiograph of the ankle is obtained with the cassette positioned along the medial aspect of the foot and ankle. The foot and ankle are placed in a weight-bearing position with the beam centered at the ankle joint. The surgeon should use this view to examine the ankle joint (Fig. 1-A). The lateral radiograph of the foot is obtained with the cassette placed along the medial aspect of the foot and ankle. The foot and ankle are placed in a weight-bearing position with the foot flat on the floor, and the beam is centered at the navicular. The surgeon should use this view to examine the foot (Fig. 1-B). The lateral ankle radiograph should make the entire foot appear abnormal (Fig. 1-A); however, on the lateral foot radiograph, the entire foot should appear normal (Fig. 1-B). Therefore, this pair of lateral radiographs demonstrates that no isolated forefoot or hindfoot deformity exists, but rather that the level of deformity is between the ankle and the foot. The lateral foot radiograph becomes the planning radiograph from which paper tracings are created. The center of rotation of the ankle joint, as demarcated by the lateral process of the talus, can be identified and marked on tracing paper, and the proposed osteotomy can be plotted1-3 (Fig. 2). Deformity correction is obtained by rotating the foot in the sagittal plane (the plane of the deformity). As the calcaneus is moved in a plantar direction, the osteotomy opens more posteriorly than anteriorly, appearing as a cornucopia3. The foot height resulting from the posterior opening of the U-osteotomy either should be comparable with the contralateral foot height or may exceed that height if additional lengthening is necessary for limb equalization. The foot height can be measured clinically (as the sphyrion height22) or on a lateral standing radiograph of the foot (as the distance between the standing surface and the tip of the medial malleolus).
Surgical Technique
The patient is positioned supine on a radiolucent table. A thigh tourniquet and ipsilateral hemisacral bump are used. A tarsal tunnel decompression is typically completed before the osteotomy to prevent tibial nerve entrapment. In addition, the Achilles tendon is lengthened through a gastrocnemius recession, gastrocnemius-soleus recession, or formal tendon lengthening to facilitate repositioning of the calcaneus. We bend a 1.8-mm Ilizarov wire into a U configuration and position it on the lateral side of the foot at the level of the proposed osteotomy to help determine the position of the incision site under fluoroscopic guidance. A modified Ollier-type incision is made on the lateral aspect of the foot at that level. The peroneal tendons and the sural nerve are identified, and the superficial peroneal nerve and extensor digitorum brevis are retracted. A subperiosteal dissection is made around the intended osteotomy level, along the lateral wall of the calcaneus, dorsal to the tuberosity of the calcaneus, anterior to the Achilles tendon insertion, and dorsal to the neck of the talus. The U-osteotomy passes across the neck of the talus, through the sinus tarsi, and under the sustentaculum tali of the calcaneus to exit posterior and dorsal in the calcaneus. The intended level of the osteotomy can be marked with multiple orthogonally inserted short 1.8-mm wires along its proposed course and checked with fluoroscopy. The osteotomy is completed by making multiple holes with use of 1.8-mm Ilizarov wires and then performing multiple passes with a small Hoke osteotome, taking care not to penetrate the soft-tissue structures on the medial side. Multiple small osteotomes can be used simultaneously to create the U-osteotomy. Once the U-osteotomy has been completed and the distal portion mobilized so that it can be translated, the tourniquet is deflated. Hemostasis is obtained, and the incision is closed. Other corrective foot and ankle procedures can be performed as needed. Next, external fixation is applied. Although the Ilizarov construct was used in four cases during the initial years of the study, our preferred type of external fixation, the Taylor Spatial Frame (Smith & Nephew, Memphis, Tennessee), was used in the remaining eleven patients.
The Taylor Spatial Frame consists of a full proximal tibial ring and a foot ring. Fixation of the proximal full ring is accomplished using 1.8-mm wires and 6-mm half-pins, and the foot ring is fixed using smooth wires with or without olive wires in the calcaneus (two wires) and the midfoot (one wire). Two smooth wires and an optional olive wire (called stirrup wires) are also placed through the talus and are attached to the tibial ring. The two smooth talar stirrup wires are inserted from medial to lateral, bent proximally by 90°, and attached to the tibial ring without tension. The optional third stirrup wire (olive wire) can be used as a third point of talar fixation to increase talar stability. The olive wire can be placed either through the plantar aspect of the sustentaculum tali (the middle facet of the subtalar joint) into the talus or obliquely from the posterolateral aspect to the anteromedial aspect of the talus.
Next, the six struts of the Taylor Spatial Frame are connected, and centering radiographs are made utilizing markers on the reference ring to calculate and calibrate the mounting parameters. The ring sizes and strut lengths and sizes are recorded, and a distal reference is chosen. These parameters are entered into an Internet-based computer program (Smith & Nephew) to generate a schedule for the external fixation adjustments. Correction is started five to seven days postoperatively. Follow-up visits are conducted every ten to fourteen days during distraction and on a monthly basis during consolidation (Figs. 1 through 7 and Appendix; all figures except Figure 2 are from the same patient).
Statistical Analysis
Two-tailed paired-sample tests were used to determine whether a significant difference existed between preoperative and final postoperative radiographic measurements. A p value of ≤0.05 was considered significant.
Source of Funding
No external funding was received for this study.
The presenting cause of the foot deformity was congenital in seven patients, traumatic in three, and developmental in five (see Appendix). The mean patient age at the time of surgery was twenty years (range, four to sixty-three years). The mean duration of external fixation was five months (range, three to eleven months). A Taylor Spatial Frame was used in eleven patients, and an Ilizarov construct was used in four. All fifteen patients achieved osseous union and were able to stand with the foot in a plantigrade position. Patients returned to wearing regular footwear at a mean of six months (range, three to eight months) postoperatively. Fourteen patients were able to walk without supports or assistance; one used only one cane or crutch to walk.
A total of seven complications occurred in four patients. These included deep pin-track infection in two, premature consolidation of the osteotomy in two, postoperative tarsal tunnel syndrome requiring decompression in two, and peroneal nerve entrapment in one. Tarsal tunnel decompression had been performed prophylactically in all patients except the two patients who required decompression following the osteotomy.
Radiographs were made and clinical examinations were performed on all fifteen patients preoperatively and at the time of final follow-up. The mean duration of clinical and radiographic follow-up was five years (range, three to nineteen years). Preoperative and postoperative radiographic measurements were obtained from axial and lateral radiographs (Table I). Comparison of the preoperative and final postoperative hindfoot alignment to evaluate the relation of the calcaneus to the tibia showed that the mean change in the calcaneotibial angle was 18° valgus (range, 6° to 40° valgus) (p = 0.003). The calcaneus was translated posteriorly by a mean of –8 mm (range, –2 to –20 mm) (p = 0.001). The foot height increased by a mean of 20 mm (range, 3 to 40 mm) (p < 0.001). The mean change in the calcaneal inclination angle was 3° (range, 0° to 19°), which was not significant (p = 0.302).
Nine of the fifteen patients completed the AOFAS questionnaire at the time of final follow-up. The mean score was 75 out of a possible 100 points. Seven patients had little or no trouble balancing, one patient had quite a bit of trouble balancing, and one had a great amount of trouble balancing. Seven of nine patients experienced no pain when performing moderate everyday activities.
Nine of the fifteen patients completed the SF-12 health survey. The mean physical and mental component summary scores were 47.9 and 52.1, respectively. The mean physical component summary score indicated that the patients who completed the survey at the time of final follow-up were not physically as healthy as an average person.
U-osteotomy with gradual correction by means of external fixation can lead to successful correction of complex foot deformity; however, the procedure is challenging and not without complications. Even in a practice specializing in deformity correction, we performed only fifteen of these procedures in a twenty-one-year period on patients with deformity resulting from a wide range of conditions. The most common deformities of the foot relative to the tibia were angulation in the frontal plane, translation in the sagittal plane, and limb-length discrepancy. Although a triple arthrodesis may appear to be an easier and more feasible treatment option, triple arthrodesis does not restore heel height or leg length, allow for correction of large-magnitude deformities, or correct complex multiplanar deformity of the foot relative to the tibia. Despite the fact that all fifteen patients had a deformity of the foot relative to the tibia at presentation, none had ankle pain and all achieved a plantigrade foot with a mean increase in foot height of 20 mm. The need for arthrodesis of the ankle joint was avoided or at least postponed.
We are aware of only six other cases of U-osteotomy reported in the literature2,4,7,8. In 1992, Ilizarov described the use of multiple osteotomies to correct complex foot deformities, including one case of U-osteotomy4. Ilizarov applied a one or two-ring tibial construct, which was then connected with hinges to a supporting foot frame. Ilizarov also inserted one wire through the posterior aspect of the talus and secured it to the tibial portion with posts.
In 1993, Paley reported on the treatment of twenty-three patients (twenty-five feet) with the Ilizarov method, including two treated with U-osteotomy7. One of the latter patients had clubfoot and a flat-top talus with equinus, varus, and supination deformities. This patient experienced premature consolidation and required a repeat osteotomy. A plantigrade foot was obtained after eight months. The second patient had juvenile rheumatoid arthritis and posttraumatic injuries with an equinus deformity. A U-osteotomy was used to correct the deformity acutely. A plantigrade foot was obtained after 8.3 months, and no complications were observed. In the present study, at least one superficial pin-track infection occurred in every patient and was treated with orally administered antibiotics. Paley stated that hinge placement during the U-osteotomy is crucial to avoid translational deformities7.
In 2005, Paley presented two additional cases of U-osteotomy2. One patient had a congenital short femur with an equinus foot and a flat-top talus that was corrected acutely with the U-osteotomy by displacement of the neck of the talus and Achilles tendon lengthening. The other patient had fibular hemimelia with a flat-top talus and a limb-length discrepancy; the U-osteotomy was used to correct both angular and length deformities. The author did not state whether complications occurred.
In 2002, Kocaoğlu et al. reported on the treatment of twenty-three deformed feet in twenty-two patients; one patient with pes planovalgus secondary to trauma underwent U-osteotomy8. The authors performed a concomitant tarsal tunnel decompression to protect the posterior tibial nerve before application of the frame. The frame was constructed in three parts: a two-ring frame on the leg, a heel ring, and a forefoot ring. Two wires were used in the mobile bone segment of the U-osteotomy. The external fixator was maintained for two months, after which the patient achieved a plantigrade foot. The patient was followed for a period of thirty-one months, and the authors reported only minor pin-track infections.
We prefer gradual correction of the U-osteotomy by means of external fixation rather than acute correction because gradual correction allows for lengthening or increasing heel height and enables gradual neurovascular stretching. Additional tension on the tibial nerve can be prevented by performing prophylactic tarsal tunnel decompression during the initial operation, and we recommend this8,23. If prophylactic decompression is not performed, recognition of the signs and symptoms of posterior tibial nerve compression is necessary and prompt decompression is imperative if it occurs24. In our study, thirteen of fifteen patients underwent prophylactic tarsal tunnel release. Both of the patients who did not undergo prophylactic tarsal tunnel release required later decompression.
Ilizarov4 and Paley2 did not state whether complications occurred. We observed two deep pin-track infections in our series, consistent with the reports by Kocaoğlu et al.8 and Paley7. Finally, premature consolidation occurred as a complication in two of our patients. Paley also observed premature consolidation in one of the two patients treated with U-osteotomy7.
Comprehension of the basic science of bone regeneration is necessary to prevent premature consolidation, especially because the calcaneal osteotomy heals quickly25-27. To avoid premature consolidation, the osteotomy surfaces should be disengaged by distraction. Also, the osteotomy surfaces should be distracted before performing angular or rotational corrections to avoid impingement of the osteotomy surfaces. Regeneration of bone occurs when these surfaces are distracted at a rate of approximately 1 mm per day2,26,27. If the rate of distraction is faster than this, bone may fail to form between the osteotomy surfaces, leading to delayed union, nonunion, or loss of alignment and length. Our correction started after five to seven days, compared with the two to seven days recommended in the literature to prevent premature consolidation1-8,28.
Another complication that could be associated with the U-osteotomy is knee flexion deformity, especially in cases in which large increases in heel height are obtained. Knee flexion contracture was not reported as a complication in any of the six cases reported in the literature2,4,7,8 and was not observed in our study. However, we recognize that tension on the Achilles tendon can produce a knee flexion deformity because the gastrocnemius muscle crosses posterior to the knee joint29.
The mean duration of treatment with external fixation was five months in our study compared with sixty-two days in the report by Ilizarov4 and eight months in the 1993 report by Paley7. The mean duration of follow-up in our study was five years; to the best of our knowledge, this is longer than the other follow-up durations reported in the literature2,4,7,8.
The present study was not prospective, and we had a relatively small sample size. Also, we lacked comparative preoperative clinical outcome data. A prospective database would better serve this inhomogeneous group of patients and allow investigators to compare the U-osteotomy with other methods of correction of complex foot deformities.
We conclude that the U-osteotomy with gradual correction by means of external fixation is a viable alternative for obtaining a plantigrade foot in cases of complex multiplanar deformity of the foot relative to the tibia. Although complications were associated with this surgical procedure, and one-half of these required a second operation, no complication precluded a successful final outcome.
Note: The authors wish to thank Amanda Chase, MA, Dori Kelly, MA, Alvien Lee, Joy Marlowe, MA, Terry-Elinor R. Reid, BS, and Stacy C. Specht, MPA, for their invaluable assistance with the manuscript.