Modified Judet Quadricepsplasty and Ilizarov Frame Application for Stiff Knee After Femur Fractures

TECHNICAL TRICK Modified Judet Quadricepsplasty and Ilizarov Frame Application for Stiff Knee After Femur Fractures Dae-Hee Lee, MD,* Tae-Ho Kim, MD, PhD, Se-Joon Jung, MD, Eun-Jong Cha, PhD, and Seong-Il Bin, MD Summary: Limited knee range of motion caused by posttraumatic arthrofibrosis after periarticular fracture remains a challenging problem. Traditionally, Judet or Thompson quadricepsplasty has been performed for severe stiff knee after distal femoral fracture. However, these procedures have a high rate of complications such as skin necrosis, wound dehiscence, and extension lag caused by long incisions and extensive surgery. The present study reports on a novel method that combines modified quadricepsplasty and the Ilizarov frame for management of severe stiff knee after metadiaphyseal fracture around the knee joints. This technique was applied in 10 patients over a 7-year period by one surgeon. The treatment resulted in improved knee range of motion without rebound phenomenon, which is a frequent problem when using either the Ilizarov frame or quadricepsplasty alone. Key Words: quadricepsplasty, extension lag, joint distraction, rebound phenomenon, stiff knee (J Orthop Trauma 2010;0:000 000) INTRODUCTION Stiff knee after periarticular fracture of the knee joint is a complex and serious complication 1 that can reduce the ability to perform daily activities. Knee stiffness is difficult to treat, and although methods of internal fixation and early rehabilitation have improved, inadequate range of motion of the knee joint after high-energy trauma may be unavoidable because of severe soft tissue injury, comminuted fractures, and multiple fractures. Among the treatment methods for stiff knee are the Thompson 2 and Judet 3 quadricepsplasty techniques and their Accepted for publication October 27, 2009. From the *Department of Orthopaedic Surgery, Korea University School of Medicine, Anam Hospital, Seoul, Korea; Department of Orthopaedic Surgery, Cheong-ju St. Mary s Hospital, College of Medicine, Catholic University, Seoul, Korea; Department of Orthopedic Surgery, University of Ulsan, College of Medicine, Asan Medical Center, Seoul, Korea; and Biomedical Engineering Department, School of Medicine, Chungbuk National University, Seoul, Korea. This work was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (2009-0063258). Reprints: Dr. Seong-Il Bin, MD, Department of Orthopaedic Surgery, Asan Medical Center, College of Medicine, Ulsan University, 388-1, Pungnap- 2-dong, Songpa-gu, Seoul 138-736, South Korea (e-mail: sibin@amc. seoul.kr). Copyright Ó 2010 by Lippincott Williams & Wilkins modifications. 4,5 Although Judet quadricepsplasty results in a lower rate of extension lag than Thomson quadricepsplasty, both methods have high rates of associated complications, including skin necrosis, wound dehiscence, and extension lag caused by a long incision and extensive surgery, as well as profound postoperative edema of the extremity and severe pain during the early postoperative period. 6 Gradual joint distraction using an Ilizarov frame, which can control the mechanical forces directly through the skeleton and allow ambulatory treatment, has been used to treat severe stiff knee. 7,8 However, removal of the Ilizarov frame can result in a loss of flexion secondary to a rebound phenomenon. 9,10 The present report describes the use of modified Judet arthroplasty in combination with Ilizarov external fixation to treat severe stiff knee after trauma. This approach led to a gain in range of knee motion and no rebound phenomenon. SURGICAL TECHNIQUE After induction of general anesthesia, passive range of motion in the affected limb was assessed and recorded. Previous incisions were identified. The limb was exsanguinated, and a tourniquet was inflated to approximately 300 mm Hg. The modified Judet s quadricepsplasty comprised three steps. These were sequential procedures with the range of knee motion determined by manipulation after each stage of dissection. Dissection was ceased at any stage of the procedure if adequate flexion/extension (usually more than 90 of flexion and less than 5 of extension) was obtained. Step 1 involved removal of intra-articular obstacles. A lateral incision was made extending from just lateral to the superior pole of the patella to the level of the distal one third of the femur (Fig. 1A). A lateral parapatellar arthrotomy was made by incising the lateral retinaculum from the superolateral corner of the patella down to the lateral aspect of its lower pole. The lateral retinaculum was freed from the lateral condyle of the femur to restore the lateral recess. Unless release of the suprapatellar pouch and medial gutter was possible, a short medial incision with a parapatellar approach was made (Fig. 1B); this approach also enabled the removal of any scar tissue encountered in the infrapatellar fat pad. Step 2 focused on removal of extraarticular obstacles between the distal femur and the quadriceps. To avoid further incisions, the medial and lateral incisions made during step 1 were used in the step 2 procedure. The tendinous portion of the vastus medialis, intermedius, and lateralis were freed from the distal one third of the femur (Fig. 2). However, the insertion sites of the vastus medialis and lateralis to the patellar tendon were not released J Orthop Trauma Volume 0, Number 0, Month 2010 www.jorthotrauma.com 1

Lee et al J Orthop Trauma Volume 0, Number 0, Month 2010 FIGURE 2. Cross-section of distal 1/3 of femur showing the three vasti muscles detached from distal femur (blue line drawn through direction of scalpel) by medial and lateral incision. FIGURE 1. Lateral (A) and medial (B) incisions were used to perform a quadricepsplasty and arthrotomy. to prevent extension lag after surgery. After retraction of the freed vastus medialis and lateralis, adhesion of the vastus intermedius to the distal femur was also released by blunt dissection, but this was not extended to the proximal two thirds of the thigh to prevent damage to the perforating artery of the vastus lateralis. The original Judet s quadricepsplasty technique included a third step, in which the lateral incision was extended proximally and anterolaterally over the hip, followed by the detachment of the origin of the rectus femoris from the anterior inferior iliac spine. However, if after steps 1 and 2 the knee range of motion was not acceptable (usually flexion less than 90 ), the third step of the original Judet s quadricepsplasty was not performed, and instead we applied Ilizarov circular fixators for gradual joint distraction and flexion using preconstructed four-level frames consisting of two tibial and femoral rings (Fig. 3A, B). A 5/8 or half ring femoral frame shape was used depending on the indication for each case. A key aspect of our frame fixation technique was the use of more half pins than wires to reduce the soft-tissue impingement, especially at the level of the distal femur. Previously, two oblique wires at this level transfixed the quadriceps mechanism and created knee extension contracture. In our method, these two wires were replaced by half pins inserted in a posteromedial and a posterolateral position, respectively. The convergence angle of the two pins was adjusted to be at least 60 degrees because the rigidity of the frame upon bending decreases markedly if the angle is less. Insertion of the transverse distal femoral wire was another important process in the frame fixation. This is a transcondylar wire inserted in the transverse line at the level of the superior pole of the patella. Great care must be taken to ensure the wire does not impinge on the soft tissue and limit knee motion. This means that the wire is inserted such that there is no movement in the wire as the knee is examined through a 0 to 90 degree range of motion. If the wire does not bend with knee movement then its position is acceptable. If it bends on both the medial and lateral sides it must be withdrawn completely and replaced. If it bends on the lateral side only, the tensor fascia lateae must be progressively divided until the wire no longer bends with knee movement. After distal femoral ring fixation, a proximal femoral ring at the level inferior to the gluteal fold was fixed using one half pin from posterolateral to anteromedial and one wire through the lateral intermuscular septum. The remaining two tibial rings were fixed approximately 7 to 8 cm distal to the knee joint and 10 to 12 cm proximal to the ankle joint. At the level between these two tibial rings, the anteromedial border or medial surface of the tibia is easily palpable, which provides a useful guide for the starting point of the half pin or Ilizarov wire. Details of pin configuration for each femoral and tibial ring are illustrated in Figure 4. The femoral and tibial rings were connected with distraction rods and medial and lateral hinges. Medial and lateral hinges were placed at the level of the knee s center of rotation. Although the center of rotation of the knee joint 2 www.jorthotrauma.com q 2010 Lippincott Williams & Wilkins

J Orthop Trauma Volume 0, Number 0, Month 2010 Judet Quadricepsplasty and Ilizarov Frame AU1 FIGURE 3. Clinical photo showing front (A) and side (B) view of entire frame with two hinges and two posterior threaded rods. varied with the knee flexion angle, we considered the desired center of rotation (isometric point) as a point on the lateral femoral condyle immediately proximal to where the popliteus and fibular collateral ligament cross one another. Lateral plane fluoroscopic radiographic images were used to concentrically align the posterior aspects of the femoral condyles, after which a point equidistant from the inferior and posterior joint space at Blumensaat s line was determined. This point usually correlates on fluoroscopy to an intersection point of the posterior cortical line of the distal femur and Blumensaat s line (Fig. 5A, B). Meticulous hemostasis was performed using electrocautery. Because we have found that postoperative hemarthrosis contributes significantly to pain and flexion contractures and may cause a more intense inflammatory response, suction drains were also routinely used to prevent hemarthrosis and were left in place for 24 to 48 hours after surgery. Compression dressings with sterilized sponges were inserted into the Ilizarov frames with the knee joint in full extension, and cotton and elastic bandages were applied from the toe to the thigh to reduce edema, swelling, and pain. POSTOPERATIVE MANAGEMENT Gradual correction of the flexion contracture began 2 days after surgery. After gradual distraction to make room for the posterior condyle, further flexion of the distal femur was achieved. A simple lateral radiograph of the injured knee joint taken 1 week after surgery was used to determine the degree of distraction, which was usually conditioned to reserve the working space for the posterior femoral condyle so that the latter did not contact the tibial articular surface during increased flexion of the knee joint. With the knee joint fully extended, gradual flexion was advanced at a rate of 3 to 4 mm per day by performing a 1-degree correction three or four times a day (Fig. 6). Early quadriceps muscle activity was advocated. Although each patient was assigned an individualized physiotherapy regimen, every patient performed passive flexion exercises for 1 hour at an interval of 4 to 6 hours daily, even during gradual distraction. For these exercises, the hexagonal and square nuts of the connecting rods and hinges between the femur and tibia were released to allow free motion of the framework. After achieving the tolerated flexion, the extension exercise was maintained with the gaining of flexion to prevent the rebound phenomenon. The hinge was locked at night in flexion or extension on an alternating nightly basis. In general, patients were taught selfassisted motion exercises, and family members also became involved to ensure that patients continued with treatment at home and on weekends. Quadriceps stretching into supine was passive, whereas hamstring exercises were active. With the Ilizarov frame hinge locked, patients were prescribed toe-touch and weight-bearing exercises and were advanced to full weight bearing by 2 weeks. The Ilizarov apparatus was removed when the patient achieved nearly full extension (#5 of flexion contracture) and flexion greater than 90 degrees and could walk without any walking aids. If the preoperative total arc of motion (TAM) was less than 25 degrees, the frame was maintained for an additional 1 or 2 weeks after achieving the goal of TAM to prevent the rebound phenomenon. q 2010 Lippincott Williams & Wilkins www.jorthotrauma.com 3

Lee et al J Orthop Trauma Volume 0, Number 0, Month 2010 FIGURE 4. Drawing showing detailed configuration of wire (straight black lines) and half pin and relationships between neurovascular structure and each ring level (we modified pictures of Catagni s atlas, 23 after receiving written permission from author and publisher). After Ilizarov frame removal, active quadriceps strengthening and further flexion were allowed and continued until the patient had good quadriceps control, straight-leg raising, no quadriceps lag, and knee flexion which had reached an acceptable plateau (usually more than 90 ). CLINICAL SERIES The study received institutional review board approval. The inclusion criteria were severe stiff knee caused by a fracture of the femur shaft or distal femur and skeletal maturity. The Ilizarov frame was indicated in patients with FIGURE 5. (A) Knee lateral view under fluoroscopy showing center of rotation; a point (red spot) equidistant from inferior (d1) and posterior (d2) articular surface at Blumensaat s line usually correlates to a point at intersection of posterior cortical line (solid black line) of distal femur and Blumensaat s line. Posterior arc of circle with radius of d1 and d2 representing distal and posterior femoral condyle (blue dotted line). (B) Postoperative lateral view showing hinge position, which nearly coincides with center of rotation under fluoroscopy intraoperatively. 4 www.jorthotrauma.com q 2010 Lippincott Williams & Wilkins

J Orthop Trauma Volume 0, Number 0, Month 2010 Judet Quadricepsplasty and Ilizarov Frame degrees (range, 85 105 ). The mean improvement in TAM was 68.5 degrees (range, 60 90 ). The TAM at the last followup was the same or greater than at the time of frame removal in all cases except one (patient 3). The average time of Ilizarov frame application was 125 (range, 102 154) days. Patients rated pain using a visual analog scale. The mean preoperative score was 5.6 (standard deviation 1.35), whereas the mean final follow-up score was 6.3 (standard deviation 0.95). Functional outcomes were assessed by analyzing the use of ambulation aids. Of the total 10 patients, 7 were walking without any support at the final follow-up. Of these seven patients, four were walking without support, and three were walking with a cane or crutch before surgery. Of the total 10 patients, 3 required cane or crutch support at the final followup. Of these three patients, one required a cane or crutch to walk before surgery, whereas the other two patients were unable to walk before surgery. Treatment resulted in improved walking status in five patients compared with preoperatively, and all patients were satisfied with the improvement in walking postsurgery. No patients developed further extensor lag or wound infection at the quadricepsplasty incision site. Pin-track infection occurred in two (20%) patients and resolved after oral antibiotics and care of the pin sites. FIGURE 6. Advancement of square nut (rightly bent arrow) along threaded distraction rod at a rate of 3 to 4 mm in compression direction (straight arrow) was converted into 1 of flexion through hinge (upper bent arrow) around knee center of rotation. severe knee arc movement limitation (usually less than 45 ), an underlying diagnosis of distal femur or femur shaft fracture, and inadequate range of motion after previous procedures (i.e., arthroscopic arthrolysis or manipulation under anesthesia). Over a 7-year period (from July 1997 to April 2004), 10 consecutive patients who met the inclusion criteria were treated using the procedure involving modified quadricepsplasty and gradual distraction using an Ilizarov apparatus. All operations were performed by the same surgeon. The mean patient age at the time of the index procedure was 39.5 (range, 23 62) years, and the mean follow-up period was 4.3 (range, 2.1 9.4) years. The average preoperative interval from the initial operation was 35.5 (range, 10 88) months. TAM was measured using a manual goniometer preoperatively, at removal of the Ilizarov frame, and at the final follow-up. Clinical results are shown in Table 1. The mean preoperative TAM was 25 degrees (range, 5 35 ), the mean TAM at the time of frame removal was 91degrees (range, 90 100 ), and the mean TAM at the last follow-up was 93.5 DISCUSSION Modern fracture management techniques, especially in intramedullary nailing, which induces rapid union, have largely achieved the goals of early motion and prevention of stiff knee in patients with fractures of the femur. 11 However, severely traumatized fractures of the femur or of the periarticular area around the knee joint precludes early mobilization and rehabilitation, resulting in stiff knee. Treatment of severe stiff knees unresponsive to manual or arthroscopic procedures remains a challenging problem. Causes of motion loss vary according to location, and treatments should therefore be appropriately tailored. The factors limiting knee flexion have been classified as intra- and extra-articular. 1,3 Intra-articular causes of flexion loss include suprapatellar adhesions as well as medial and lateral gutter adhesions. Extra-articular causes of flexion loss include fibrosis and shortening of the lateral expansions of the vasti and their adherence to the femoral condyles, fibrosis of the vastus intermedius, and shortening of the rectus femoris. In addition, fracture callus and adherence of the skin to underlying muscle can be causes. Quadricepsplasty is traditionally classified as being either Thomson or Judet, despite the many modifications. In the Thompson technique, the vastus medialis, vastus lateralis, and vastus intermedius are freed from the rectus femoris through an anterior midline incision. The rectus femoris is then isolated from the rest of the quadriceps mechanism. This method has been associated with several complications, including delayed wound healing, skin necrosis, and extension lag. 4,12 Flexion of the knee causes tension to the wound, and pressure from the patella and femoral condyles may cause ischemia of the skin, thus giving rise to delayed wound healing and skin necrosis. A more serious problem related to the q 2010 Lippincott Williams & Wilkins www.jorthotrauma.com 5

Lee et al J Orthop Trauma Volume 0, Number 0, Month 2010 TABLE 1. Preoperative, Postoperative, and Final Follow-Up Results After Judet s Quadricepsplasty With Ilizarov Frame Application Preoperative ROM (degrees) Intraoperative ROM (degrees) Case Sex Age (yr) Side FC FF TAM FC FF TAM 1 M 30 L 5 40 35 0 75 75 2 M 47 L 20 55 35 0 85 85 3* M 47 L 0 15 15 0 75 75 4 M 23 R 20 30 10 5 70 65 5 M 40 R 5 30 25 0 70 70 6 M 44 L 15 45 30 0 80 80 7 F 32 L 30 35 5 5 85 80 8 M 24 R 10 45 35 0 70 70 9 F 62 R 0 30 30 0 80 80 10 M 30 R 0 30 30 0 75 75 ROM after Removal of Ilizarov (degrees) Extension Lag Final Follow-Up ROM (degrees) Case FC FF TAM Preoperative Postoperative FC FF TAM Improvement in TAM Preoperative VAS Postoperative VAS 1 0 90 90 5 5 0 95 95 60 6 6 2 0 95 95 0 0 0 95 95 60 8 8 3* 0 95 95 0 0 0 85 85 70 5 7 4 5 95 90 0 0 0 90 90 80 5 7 5 0 90 90 0 0 0 90 90 65 5 6 6 0 95 95 0 0 0 95 95 65 7 7 7 5 95 90 0 0 5 100 95 90 4 5 8 0 90 90 10 10 5 110 105 70 5 6 9 0 100 100 0 0 0 100 100 70 7 6 10 0 95 95 0 0 0 95 95 65 4 5 Case numbers are presented in chronological order according to when surgery was performed. *Case showing reduction in total arc of motion between Ilizarov removal and final follow-up. FC = flexion contracture; FF = further flexion; ROM = range of motion; TAM = total arc of motion; VAS = visual analogue scale. biomechanics of the knee joint is extension lag, which, if permanent, may affect stability of the knee. Disruption of the quadriceps muscle can result in a weakened extensor mechanism with extension lag. 6,13,14 Extension lag appears to occur less commonly after the Judet compared with the Thompson technique. However, this complication can still occur in the Judet approach, 15 even in minimally invasive salvage procedures, because of surgical insult to the extensor mechanism. 16 Another disadvantage of the original Judet quadricepsplasty technique is the need to make a long lateral incision, which may damage the perforating vessels of the vastus lateralis. These vessels enter the muscle 7 to 8 cm from the greater trochanter 17 and may give rise to a hematoma unless meticulous hemostasis is performed. Moreover, the large incision may cause a profound inflammatory reaction, especially during hematoma formation, resulting in postoperative edema in the extremity and severe pain during the early postoperative period. 16 Therefore, in the present series, we did not perform step 3 of the original Judet quadricepsplasty technique, which involves detachment of the origin of the rectus femoris in the anterior inferior iliac spine. In addition, we did not extend the incision into the proximal two thirds of the thigh to prevent damage to the perforators of the vastus lateralis and severe edema. As an alternative to step 3, we performed mechanical distraction using the Ilizarov apparatus. The consequent gradual and continuous correction not only induces stretching but real growth of the shortened soft tissues. 18 Ilizarov joint distraction can gradually loosen intraarticular and extraarticular fibrous tissue in stiff joints and, if necessary, can be combined with gradual relocation of the bones to their correct joint positions. 8 In addition, this gradual distraction method prevents refracture caused by avulsion of the surrounding soft tissue during rehabilitation to gain further flexion and minimizes the risks of neurovascular complications. One disadvantage of the Ilizarov procedure is the rebound phenomenon, in which flexion contracture recurs after the removal of the apparatus. 10,15,19,20 However, of our 10 patients who underwent this procedure, only 1 showed evidence of the rebound phenomenon. Although the precise reason for this low rebound phenomenon frequency remains unknown, the amount of soft-tissue release in the combined quadricepsplasty may compensate for the recontracture of soft tissue after removal of the Ilizarov apparatus. In addition, periods of passive stretch for a constant time every day may be sufficient to prevent loss of sarcomeres and to maintain normal muscle connective tissue proportions. 21 This stretch may not only result in elevated stretch tolerance but also viscoelastic accommodation of the muscle fibers. Therefore, the achieved range of motion is not transient, and the muscle is unlikely to 6 www.jorthotrauma.com q 2010 Lippincott Williams & Wilkins

J Orthop Trauma Volume 0, Number 0, Month 2010 Judet Quadricepsplasty and Ilizarov Frame return to its preoperative characteristics. 22 During hospitalization, our patients underwent as much passive stretch rehabilitation as possible each day, and patients and their families were encouraged to continue these passive and active exercises at home even after removal of the Ilizarov fixator. We believe that this vigorous and continuous rehabilitation program is important for a successful outcome and reduces the risk of the rebound phenomenon occurring. We found that patients reported slightly greater pain after surgery. A change from stable to mobile joints can lead to increased surface contact between the tibiofemoral joints, and chondral damage caused by the initial trauma may cause greater postoperative pain. It is therefore important to inform patients that although surgery may lead to an increased range of motion, it may also result in increased pain. The present treatment involving modified quadricepsplasty and an Ilizarov external fixator can be considered a successful salvage procedure for severely stiff knees after femur fractures that fail less invasive techniques. Furthermore, this procedure was associated with low incidences of extension lag after preexisting quadricepsplasty and rebound phenomenon after isolated gradual distraction using the Ilizarov apparatus. ACKNOWLEDGMENT The authors thank Professor Maurizio A. Catagni, MD, from the Medical School of the University of Milan, for kindly granting permission to use pictures from his book. REFERENCES 1. Nicoll EA. Quadricepsplasty. J Bone Joint Surg Br. 1963;45:483 490. 2. Thompson TC. Quadricepsplasty to improve knee function. J Bone Joint Surg Am. 1944;26:366 379. 3. Judet R. Mobilisation of the stiff knee. Proceedings of the British Orthopaedic Association. J Bone Joint Surg Br. 1959;41:856 857. 4. Ebraheim NA, DeTroye RJ, Saddemi SR. Results of Judet quadricepsplasty. J Orthop Trauma. 1993;7:327 330. 5. Hahn SB, Lee WS, Han DY. A modified Thompson quadricepsplasty for the stiff knee. J Bone Joint Surg Br. 2000;82:992 995. 6. Moore TJ, Harwin C, Green SA, et al. The results of quadricepsplasty on knee motion following femoral fractures. J Trauma. 1987;27:49 51. 7. Devalia KL, Fernandes JA, Moras P, et al. Joint distraction and reconstruction in complex knee contractures. J Pediatr Orthop. 2007; 27:402 407. 8. van Roermund PM, van Valburg AA, Duivemann E, et al. Function of stiff joints may be restored by Ilizarov joint distraction. Clin Orthop Relat Res. 1998;348:220 227. 9. Bar-Meir E, Yaffe B, Winkler E, et al. Combined Ilizarov and free flap for severe recurrent flexion-contracture release. J Burn Care Res. 2006;27: 529 534. 10. Herzenberg JE, Davis JR, Paley D, et al. Mechanical distraction for treatment of severe knee flexion contractures. Clin Orthop Relat Res. 1994;301:80 88. 11. Hansen ST, Winquist RA. Closed intramedullary nailing of the femur. Küntscher technique with reaming. Clin Orthop Relat Res. 1979;138: 56 61. 12. Warner JJ. The Judet quadricepsplasty for management of severe posttraumatic extension contracture of the knee. A report of a bilateral case and review of the literature. Clin Orthop Relat Res. 1990;256: 169 173. 13. Pick RY. Quadricepsplasty: a review, case presentations, and discussion. Clin Orthop Relat Res. 1976;120:138 142. 14. Ratliff AH. Quadricepsplasty. Injury. 1972;4:126 130. 15. Merchan EC, Myong C. Quadricepsplasty: the Judet technique and results of 21 posttraumatic cases. Orthopedics. 1992;15:1081 1085. 16. Wang JH, Zhao JZ, He YH. A new treatment strategy for severe arthrofibrosis of the knee. A review of twenty-two cases. J Bone Joint Surg Am. 2006;88:1245 1250. 17. Alici T, Bulucx L, Tosun B, et al. Modified Judet s quadricepsplasty for loss of knee flexion. Knee. 2006;13:280 283. 18. Fink B, Schwinger G, Singer J, et al. Biomechanical properties of tendons during lower-leg lengthening in dogs using the Ilizarov method. J Biomech. 1999;32:763 768. 19. Huang SC. Soft tissue contractures of the knee or ankle treated by the Ilizarov technique. High recurrence rate in 26 patients followed for 3 6 years. Acta Orthop Scand. 1996;67:443 449. 20. Damsin JP, Ghanem I. Treatment of severe flexion deformity of the knee in children and adolescents using the Ilizarov technique. J Bone Joint Surg Br. 1996;78:140 144. 21. Williams PE. Effect of intermittent stretch on immobilised muscle. Ann Rheum Dis. 1988;47:1014 1016. 22. Magnusson SP. Passive properties of human skeletal muscle during stretch maneuvers. A review. Scand J Med Sci Sports. 1998;8:65 77. 23. Catagni MA. In: Bianchi Maiocchi A, ed. Atlas for the Insertion of Transosseous Wires and Half-Pins Ilizarov Method. 2nd ed (revised). Milan, Italy: Medi Surgical Video, 2003. q 2010 Lippincott Williams & Wilkins www.jorthotrauma.com 7