Browse the corpus

Knee rotation was not changed postoperatively in the UG, as expected. In the BG, knee rotation changed by as much as one-fourth of FDO, even though theoretically it should not be affected by the FDO. Transverse kinematic data of the BG were confusing as the improvement in FPA took place at the femur and tibia. This might be because of measuring error that is prone to occur in the transverse plane compared with other planes. There are still debates on whether compensatory pelvic rotation is corrected after FDO. Some authors, including Chung et al. 15 and Christopher et al. 16, reported that pelvic rotation in spastic diplegia was not changed even after normalization of femoral torsion by FDO. There are more debates on spastic hemiplegias. Gage et al. 6 reported that pelvic external rotation cannot be corrected after FDO in hemiplegic patients. He recommended a slight undercorrection during FDO for hemiplegias because of the above-mentioned reason. Meanwhile, Christopher et al. 16 reported that hemiplegic patients showed a significant improvement in pelvic rotation after FDO. In this study, the overall pelvic rotation was not changed after FDO in both groups.

Introduction Acute elbow extension deficit is an unusual phenomenon that has been observed in patients with congenital radioulnar synostosis. This deficit has been attributed to the presence of a tight, hypoplastic annular ligament that traps the radial head 1–5. We report the case of an 11-year-old girl with previously undiagnosed congenital radioulnar synostosis, who developed an acute extension deficit of the right elbow after a fall. The elbow extension deficit was restored following lateral capsular release. Case report An 11-year-old, right-handed, otherwise healthy girl suffered a fall from standing and subsequently developed pain and limited range of movement (ROM) in her right elbow. Initial radiographs revealed congenital radioulnar synostosis with anterior dislocation of the radial head (Fig. 1). The patient and her family reported no history of pain, stiffness, or decreased ROM involving the right elbow before this injury. The patient underwent closed manipulation. Examination under anesthesia revealed a ROM from 0° to 135°. Snapping occurred as the elbow was extended from a position of maximal flexion to 30° but did not occur from extension back to flexion. In addition, the right forearm was fixed in 5° of supination. Fig. 1 Preoperative right elbow lateral radiograph. A radioulnar synostosis and anterior radial head dislocation are present.

Case report An 11-year-old, right-handed, otherwise healthy girl suffered a fall from standing and subsequently developed pain and limited range of movement (ROM) in her right elbow. Initial radiographs revealed congenital radioulnar synostosis with anterior dislocation of the radial head (Fig. 1). The patient and her family reported no history of pain, stiffness, or decreased ROM involving the right elbow before this injury. The patient underwent closed manipulation. Examination under anesthesia revealed a ROM from 0° to 135°. Snapping occurred as the elbow was extended from a position of maximal flexion to 30° but did not occur from extension back to flexion. In addition, the right forearm was fixed in 5° of supination. Fig. 1 Preoperative right elbow lateral radiograph. A radioulnar synostosis and anterior radial head dislocation are present. After awakening from anesthesia, the patient could flex her elbow to 135° but could extend her elbow only to 30°. The right forearm remained fixed in 5° of supination. Treatment with physical therapy and anti-inflammatory drugs for 1 month failed to restore her ROM to that observed under anesthesia. MRI of the right elbow revealed a bony fusion between the proximal one-third of the ulna and the radius, involving the radial head and the coronoid process. The radial head was displaced anteriorly, placing tension on the anterior capsule. Alignment of the olecranon–humeral articulation was normal. TheT2-weighted sequences showed an effusion within the elbow joint and edema in the anterior soft tissues (Fig. 2). On the basis of the MRI results, a diagnosis of type IV radioulnar synostosis 6 was established, but the reason for the patient’s limited ROM remained unclear.

e olecranon–humeral articulation was normal. TheT2-weighted sequences showed an effusion within the elbow joint and edema in the anterior soft tissues (Fig. 2). On the basis of the MRI results, a diagnosis of type IV radioulnar synostosis 6 was established, but the reason for the patient’s limited ROM remained unclear. Fig. 2 MRI of the right elbow (T2 image) demonstrates fusion of the proximal one-third of the ulna and radius. Note the anterior dislocation of the radial head, joint fluid effusion, tension of anterior capsule, and edema signal in the anterior soft tissue.

e olecranon–humeral articulation was normal. TheT2-weighted sequences showed an effusion within the elbow joint and edema in the anterior soft tissues (Fig. 2). On the basis of the MRI results, a diagnosis of type IV radioulnar synostosis 6 was established, but the reason for the patient’s limited ROM remained unclear. Fig. 2 MRI of the right elbow (T2 image) demonstrates fusion of the proximal one-third of the ulna and radius. Note the anterior dislocation of the radial head, joint fluid effusion, tension of anterior capsule, and edema signal in the anterior soft tissue. A second examination under anesthesia revealed the same snapping associated with extension that was observed during the initial examination. The elbow was then explored using a lateral approach. Overgrowth of the radial head and a concave deformity at the base of the radial neck (Fig. 3) were observed. As the elbow was flexed, the radial neck passed anterior to the supracondylar groove of the humerus. An annular ligament-like structure overlapped the radial neck and fixed it to the anterior capsule; however, there was no trapping of the radial head during flexion or extension (video, Supplemental digital content 1, http://links.lww.com/JPOB/A8). The radial head and neck articulated with the convex shape of the lateral condyle. Following the capsular incision, no locking or snapping was observed as the elbow moved from maximal flexion to maximal extension. However, following repair of the lateral capsule, locking and snapping reappeared as the elbow was extended from full flexion. The snapping was linked to the passage of the radial head over the prominence of the lateral condyle. The complete capsular repair was then undone, and a partial repair was performed with two stitches (Fig. 4). Following this revision of the capsular closure, the right elbow showed full extension and flexion without snapping. Postoperatively, the patient retained full elbow extension and flexion without snapping. A 5-year follow-up examination revealed full extension and flexion with no recurrence of the snapping. However, the right forearm was fixed in 5° of supination.

e, the right elbow showed full extension and flexion without snapping. Postoperatively, the patient retained full elbow extension and flexion without snapping. A 5-year follow-up examination revealed full extension and flexion with no recurrence of the snapping. However, the right forearm was fixed in 5° of supination. Fig. 3 Lateral exposure of the right elbow. The lateral collateral ligament was incised and opened. The base of the radial neck (a) had concave deformations. An annular ligament-like structure (b) overlapped the radial neck and fixed it to the anterior capsule and the lateral collateral ligament. The radial head (c) was seated on the anterior aspect of the distal humerus. A supracondylar groove was observed concave to the cartilage surface (d). While extending the elbow to 30° from full flexion, the dislocated radial head was observed to pass over the prominence of the lateral condyle (e). Fig. 4 Lateral exposure of the right elbow: the complete capsular repair was then removed and partial repair was performed with two stitches (a).Video, Supplemental digital content 1, http://links.lww.com/JPOB/A8: Snapping was observed during the extension from full flexion to 30°, but not during flexion from full extension. This snapping was no longer observed following lateral release.

ir was then removed and partial repair was performed with two stitches (a).Video, Supplemental digital content 1, http://links.lww.com/JPOB/A8: Snapping was observed during the extension from full flexion to 30°, but not during flexion from full extension. This snapping was no longer observed following lateral release. Discussion Proximal radioulnar synostosis is a rare congenital upper-extremity disorder that can be associated with a 10–30° limitation in terminal elbow extension 7. A literature review identified 16 case reports on elbow motion deficits occurring in patients with type IV radioulnar synostosis, and a summary comparison between these reports is presented in Table 1. Table 1 Reported cases of elbow motion deficit in patients with type IV congenital proximal radioulnar synostosis 6 Unlike the present case, no history of elbow trauma was documented before the development of the elbow extension deficit in the earlier published case reports. The indication of surgery in general is to improve extension deficit in this kind of congenital anomaly. Elbow extension deficit to some extent does not interfere with normal function in daily life. The major function of the upper extremity is open kinetic chain movement, where the hand is free to move, such as eating, throwing, carrying objects. However, some weight-bearing activities of the upper extremity (push-up, falling down protection by hand, and elbow extension) may need full extension of the elbow to increase joint stability.

the upper extremity is open kinetic chain movement, where the hand is free to move, such as eating, throwing, carrying objects. However, some weight-bearing activities of the upper extremity (push-up, falling down protection by hand, and elbow extension) may need full extension of the elbow to increase joint stability. The different surgical methods were compared, and only in four of the 16 cases was the snapping problem solved solely by excision of annular ligament-like tissue (ALLT). Shinohara et al. 3 reported on two arthroscopic cases that required more advanced techniques for the excision of this structure to restore the elbow ROM. Masuko et al. 2 also proposed the lateral approach to excise ALLT, with good results through intensive exploration. With the exception of the excision of the ALLT, other findings at the time of surgery included hypertrophy of the anterior capsule, which required treatment with capsular excision 4,8. Most cases (10/16) show overgrowth of the radial head, which is treated with excision 5,9,10, and the radial head was usually entrapped by hypertrophic ligamentous tissues in the flexion position. However, radial head resection may affect the weight-bearing ability of the upper extremity and compromise elbow function on the basis of anatomical and biomechanical considerations.

h is treated with excision 5,9,10, and the radial head was usually entrapped by hypertrophic ligamentous tissues in the flexion position. However, radial head resection may affect the weight-bearing ability of the upper extremity and compromise elbow function on the basis of anatomical and biomechanical considerations. Lateral release is a simple method with a short operation time and less tissue damage. Moreover, a large amount of musculotendinous units originate from the lateral condyle to provide stability over the lateral aspect of the elbow joint. If the snapping persists after the lateral release method, it could be treated using an approach like ALLT or even radial head resection. Meanwhile, ultrasonography may be useful for yielding real-time information in a minor operation theater setting. In our case, tension in the anterior capsule, resulting from the progressive anomalous development of the dislocated radial head, may have contributed to the decreased elbow ROM and snapping symptoms. The minor trauma preceding the patient’s presentation likely induced joint swelling and thereby also contributed to the acute decrease in ROM. As the elbow was extended to 30° from full flexion, the radial head passed over the prominence of the lateral condyle. This motion placed additional tension on the anterior capsule, which compressed the radial head against the concavity of the lateral condyle and prevented full extension of the elbow.

decrease in ROM. As the elbow was extended to 30° from full flexion, the radial head passed over the prominence of the lateral condyle. This motion placed additional tension on the anterior capsule, which compressed the radial head against the concavity of the lateral condyle and prevented full extension of the elbow. Similar to the other reported cases, our patient was also found at surgery to have a tight and hypoplastic annular ligament. Full extension and flexion without any snapping was observed following the release of the lateral capsule. These findings supportedour proposed explanation, indicating that anterior capsular tension in this patient and the lateral capsular release had basically increased the volume of the joint to accommodate the overgrown radial head articulating with the convexity of the lateral condyle. However, the incongruent radiocapitellar articulation of the elbow remains as such. Thus, the absent forearm rotations may be due to the synostosis. In summary, lateral capsular release may be an alternative and effective surgical treatment method for acute elbow extension deficits associated with type IV radioulnar synostosis. Conclusion We described a patient with an elbow extension deficit and type IV congenital radioulnar synostosis. Her elbow flexion–extension was restored and elbow snapping was eliminated after isolated lateral capsular release. We thereby suggest that lateral capsular release be considered as a potential treatment method for acute elbow extension deficits with type IV radioulnar synostosis.

e IV congenital radioulnar synostosis. Her elbow flexion–extension was restored and elbow snapping was eliminated after isolated lateral capsular release. We thereby suggest that lateral capsular release be considered as a potential treatment method for acute elbow extension deficits with type IV radioulnar synostosis. Supplementary Material SUPPLEMENTARY MATERIAL Acknowledgements Conflicts of interest There are no conflicts of interest. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's website (http://www.jpo-b.com).

Introduction Nontraumatic sacral fractures can be characterized as insufficiency fractures in elderly populations and fatigue fractures in younger populations and athletes 1–3. Insufficiency fractures occur in osteoporotic patients or in patients who have undergone radiation therapy for pelvic malignancy. Fatigue fractures occur in physically active individuals, such as long-distance runners, in whom repetitive stress loads occur 4. Several reports have described their occurrence in athletes who engage in vigorous exercise. However, sacral fatigue fractures in children are very uncommon 4–6, especially in those younger than 10 years of age. Here, we present two cases of 9-year-old children diagnosed with sacral fatigue fractures. Case reports The authors obtained informed written consent from the patients’ parents to publish this case report. Case 1 A 9-year-old baseball player presented with a 1-month history of right low back and buttock pain during baseball. He had just returned to baseball after cessation of sports activity because of osteochondritis dissecans of the left knee. MRI showed signal changes in the right sacral ala. He was referred to our department for further workup of his low back and buttock pain.

history of right low back and buttock pain during baseball. He had just returned to baseball after cessation of sports activity because of osteochondritis dissecans of the left knee. MRI showed signal changes in the right sacral ala. He was referred to our department for further workup of his low back and buttock pain. The patient had no paresthesia, weakness, fever, bowel or bladder dysfunction, or weight loss. He had no history of previous pelvic disease, and family history was nonspecific. No history of trauma was reported. Physical examination indicated local tenderness over the right iliac crest and slight limitation in the range of motion in the right hip; however, no areas of decreased sensation or leg-length discrepancy were noted. Pain was exacerbated by activity and relieved on rest.

nonspecific. No history of trauma was reported. Physical examination indicated local tenderness over the right iliac crest and slight limitation in the range of motion in the right hip; however, no areas of decreased sensation or leg-length discrepancy were noted. Pain was exacerbated by activity and relieved on rest. Plain radiographs of the pelvis showed no abnormal findings such as osteolytic or sclerotic changes (Fig. 1). MRI showed a linear signal void in the right S1 ala on both T1-weighted and short tau inversion recovery (STIR) coronal images. Further, there were low signal intensity changes on T1-weighted images and high signal changes on STIR surrounding the linear signal void (Fig. 2a and b). A linear sclerotic change was observed at the same area on computed tomography, and spina bifida occulta (SBO) was noted from S1 to S5 (Fig. 3). A diagnosis of sacral fatigue fracture was made on the basis of both clinical and imaging data. We advised the patient to withdraw from all sports activities for 6 weeks. Three months after symptom onset, pain was improved and signal changes in the right S1 ala had disappeared on MRI (Fig. 2c and d). Fig. 1 Unremarkable initial anterior plain radiograph of the pelvis.

Plain radiographs of the pelvis showed no abnormal findings such as osteolytic or sclerotic changes (Fig. 1). MRI showed a linear signal void in the right S1 ala on both T1-weighted and short tau inversion recovery (STIR) coronal images. Further, there were low signal intensity changes on T1-weighted images and high signal changes on STIR surrounding the linear signal void (Fig. 2a and b). A linear sclerotic change was observed at the same area on computed tomography, and spina bifida occulta (SBO) was noted from S1 to S5 (Fig. 3). A diagnosis of sacral fatigue fracture was made on the basis of both clinical and imaging data. We advised the patient to withdraw from all sports activities for 6 weeks. Three months after symptom onset, pain was improved and signal changes in the right S1 ala had disappeared on MRI (Fig. 2c and d). Fig. 1 Unremarkable initial anterior plain radiograph of the pelvis. Fig. 2 Coronal MR images in case 1. (a, b) Initial examination. (c, d) Three months later. Coronal images showing (a) diffuse low marrow signals with a central linear signal void (arrows) in the T1-weighted image and (b) diffuse increased signals in the STIR image at the right S1 ala (arrows). (c, d) The low signals in the T1-weighted image, high signals in the STIR image, and the linear signal void are no longer evident 3 months later. STIR, short tau inversion recovery.

ral linear signal void (arrows) in the T1-weighted image and (b) diffuse increased signals in the STIR image at the right S1 ala (arrows). (c, d) The low signals in the T1-weighted image, high signals in the STIR image, and the linear signal void are no longer evident 3 months later. STIR, short tau inversion recovery. Fig. 3 (a) Coronal and (b) axial CT images showing linear medullar sclerosis in the right S1 ala (arrows) and SBO (arrowhead). (c) Three-dimensional-CT image showing SBO from S1 to S5. CT, computed tomography; SBO, spina bifida occulta. Case 2 A 9-year-old male baseball player experienced pain in the left hip after playing baseball. He stopped playing baseball and rested following onset. No history of paresthesia, weakness, fever, bowel or bladder dysfunction, or weight loss was noted. His previous medical history and family history were nonspecific, and there was no history of trauma. MRI showed signal intensity changes in the left sacral ala. Three weeks after symptom onset, he was referred to our department for further workup of the persistent left hip pain. However, the pain had already improved upon presentation to us. Physical examination indicated no tenderness over the paraspinal region, buttocks, low back, or femoral head. Both his lower extremities showed full range of motion and full motor strength, and there were no areas of decreased sensation or leg-length discrepancy. The FABER-Patrick test was negative on both sides. His reflexes were symmetric, bilateral, and normal. He could stand on one leg without pain.

back, or femoral head. Both his lower extremities showed full range of motion and full motor strength, and there were no areas of decreased sensation or leg-length discrepancy. The FABER-Patrick test was negative on both sides. His reflexes were symmetric, bilateral, and normal. He could stand on one leg without pain. Plain radiographs of the pelvis were normal. MRI showed a linear signal void on the T2-weighted coronal image in the left S2 sacral ala (Fig. 4a). Computed tomography indicated a linear band of medullary sclerosis at the same site and SBO was noted at S2 (Fig. 5). This patient had a transitional vertebra in the lumbosacral region (Fig. 5c). The diagnosis of sacral fatigue fracture was made. Fig. 4 T2-weighted coronal MR image in case 2. (a) Initial examination. (b) Two months later. Coronal MR images showing the linear signal void (arrows) in the T2-weighted image (a). The linear signal void in the T2-weighted image is no longer evident 2 months later (b). Fig. 5 (a) Coronal and (b) axial CT images showing linear medullar sclerosis in the left S2 ala (arrows) and SBO (arrowhead). (c) Three-dimensional-CT image showing SBO at S2. CT, computed tomography; SBO, spina bifida occulta. Because his pain had already improved, he was allowed to resume sports activities as long as pain was tolerable. Two months later, T2-weighted coronal image showed the linear signal void at the fracture to be unclear and diffuse (Fig. 4b). The signal intensity of the surrounding medulla was essentially normal. The patient had no recurrent pain or complaints and returned to sports activities.

ivities as long as pain was tolerable. Two months later, T2-weighted coronal image showed the linear signal void at the fracture to be unclear and diffuse (Fig. 4b). The signal intensity of the surrounding medulla was essentially normal. The patient had no recurrent pain or complaints and returned to sports activities. Discussion Fatigue fractures in children usually occur in the lower extremities, most commonly in the tibia, fibula, or metatarsals 7,8. Fatigue fractures of the sacrum are less commonly reported, but may be encountered as the result of serious athletic training for long-distance running 9. To our knowledge, there are only five case reports of fatigue fractures of the sacrum in children aged 12 years or younger in the literature. Sacral fatigue fractures are very uncommon in children and all published reports describe excellent clinical outcomes with conservative treatment (Table 1) 5,7,10–12. Table 1 Previous studies on sacral fatigue fracture in children aged 12 years or younger

Discussion Fatigue fractures in children usually occur in the lower extremities, most commonly in the tibia, fibula, or metatarsals 7,8. Fatigue fractures of the sacrum are less commonly reported, but may be encountered as the result of serious athletic training for long-distance running 9. To our knowledge, there are only five case reports of fatigue fractures of the sacrum in children aged 12 years or younger in the literature. Sacral fatigue fractures are very uncommon in children and all published reports describe excellent clinical outcomes with conservative treatment (Table 1) 5,7,10–12. Table 1 Previous studies on sacral fatigue fracture in children aged 12 years or younger Symptoms of sacral fatigue fractures are low back pain of gradual onset, occasionally radiating to the hip or the groin. Most patients have normal results on neurological examination, and most show full range of motion of the back and lower extremities. With respect to radiological findings, plain radiographs are usually normal as they are not sensitive enough to show sacral fatigue fractures, but may be useful in revealing other causes of back pain. Bone scintigraphy and MRI are effective for diagnosis 1,2,4–7,9–12. Bone scintigraphy shows increased activity in the lateral part of the sacrum and MRI shows bone marrow edema as an early sign of a stress fracture. T1-weighted and T2-weighted and STIR images show a linear signal void with surrounding diffuse decreased marrow signals on the T1-weighted images and surrounding high-intensity signals on the T2-weighted and STIR images. The radiological findings in our cases were consistent with these features.

n of a stress fracture. T1-weighted and T2-weighted and STIR images show a linear signal void with surrounding diffuse decreased marrow signals on the T1-weighted images and surrounding high-intensity signals on the T2-weighted and STIR images. The radiological findings in our cases were consistent with these features. In both our patients, SBO was found at each affected level of the sacral fatigue fractures. However, we could find no reports describing the presence of SBO in pediatric patients with sacral fatigue fractures (Table 1) 5,7,10–12. In addition, athletes typically incur fatigue fractures, but neither patient was a highly enthusiastic athlete. Sakai et al. 13 reported a significantly higher incidence of lumbar spondylolysis among patients with SBO than in those without SBO (16.2 vs. 5.0%, respectively. Several studies in the literature have described a positive association between SBO and spondylolysis 14–16. A recent finite element study showed a bifid arch in SBO and following laminectomy increases load across the isthmus, predisposing toward early fatigue fractures of the isthmus 17. Similarly, SBO in the sacrum might affect the stress concentration on the sacral alae, predisposing toward stress fractures of the sacrum without repetitive stress.

tudy showed a bifid arch in SBO and following laminectomy increases load across the isthmus, predisposing toward early fatigue fractures of the isthmus 17. Similarly, SBO in the sacrum might affect the stress concentration on the sacral alae, predisposing toward stress fractures of the sacrum without repetitive stress. Conclusion We have presented two cases of sacral fatigue fractures in 9-year-old patients with SBO at each affected level. Sacral fatigue fracture is an important consideration in the differential diagnosis of lower back and pelvic pain in children and should be considered whenever a healthy, active child presents with unexplained persistent low back and buttock pain. Acknowledgements Conflicts of interest There are no conflicts of interest.

Introduction The medial approach for open reduction of developmental dysplasia of the hip was first described by Ludloff in 1908 1. In 1973, Ferguson modified this approach by using a more posterior interval 2. Open reduction through a medial approach has been used in children younger than 24 months of age 2. This method has been described as a simple and atraumatic procedure that enables bloodless access to the medial and inferior sections of the hip joint 2–4. The pericapsular acetabuloplasty was described by Pemberton in 1965 5. This is a reshaping incomplete iliac osteotomy in which the acetabular roof is hinged on the triradiate cartilage. Thus, the lower fragment of the ilium is rotated forward, downward and outward 5. Pemberton acetabuloplasty (PA) enables greater correction of acetabular dysplasia by providing further lateral and anterior acetabular coverage than the Salter osteotomy 6. In our centre, the Ferguson medial approach has been applied as a first treatment option for dislocations presenting after six months of age or where splintage has failed. PA was administered in patients older than 15 months of age who had an acetabular angle greater than 35° on preoperative radiographs or instability at the end of medial open reduction. In this retrospective study, we present our experience with open reduction through a medial approach and PA in a single session in patients older than 15 months of age. To our knowledge, this is the first report of a combination of Ferguson medial open reduction with PA in a single-stage operation.

In our centre, the Ferguson medial approach has been applied as a first treatment option for dislocations presenting after six months of age or where splintage has failed. PA was administered in patients older than 15 months of age who had an acetabular angle greater than 35° on preoperative radiographs or instability at the end of medial open reduction. In this retrospective study, we present our experience with open reduction through a medial approach and PA in a single session in patients older than 15 months of age. To our knowledge, this is the first report of a combination of Ferguson medial open reduction with PA in a single-stage operation. Patients and methods We retrospectively reviewed the records of 22 patients (32 hips) who underwent a single-stage open reduction through the medial approach and PA for developmental dysplasia of the hip between 1995 and 2007. Neuromuscular or teratologic hip dislocations were not included in the study. This study was approved by the institutional ethics committee. No patient had received any previous treatment. There were 21 girls and one boy. Both hips were affected in 10 patients, the right hip in 15 patients and the left hip in 17. In patients who had bilaterally dislocated hips, bilateral medial open reduction and bilateral PA were performed during the same session. PA was performed in 26 hips that had an acetabular angle greater than 35° on preoperative radiographs and six hips that had instability at the end of medial open reduction. Femoral shortening osteotomy was not performed in any patient. All operations were performed by the senior author (C.B.). The mean age of the patients at the time of the operation was 19.8 months (16–24). The mean follow-up period was 10.9 years (7–19).

six hips that had instability at the end of medial open reduction. Femoral shortening osteotomy was not performed in any patient. All operations were performed by the senior author (C.B.). The mean age of the patients at the time of the operation was 19.8 months (16–24). The mean follow-up period was 10.9 years (7–19). Operative technique A medial approach was applied through the interval between the adductor longus and the gracilis. The lesser trochanter was identified with blunt dissection and iliopsoas tenotomy was performed. The joint capsule was carefully opened inferomedially and great care was taken to avoid damage to the medial femoral vessels. In all patients, the ligamentum teres and the pulvinar were removed. After the open reduction, a PA was performed through an anterior iliofemoral approach without capsular dissection and capsulorrhaphy. A curved cut was made on the medial and lateral walls of the acetabulum and extended down to the posterior wing of the triradiate cartilage. After the osteotomy, the lower fragment was rotated anterolaterally. A triangular bone graft was taken from the ipsilateral iliac bone and placed into the osteotomy site. No internal fixation was used. The stability of the hip was checked in different positions. Then, a hip spica cast was applied at 20° of flexion, 30° of abduction and 10° of internal rotation for 6 weeks. After the removal of the cast, an abduction brace was used at all times for 6 weeks.

laced into the osteotomy site. No internal fixation was used. The stability of the hip was checked in different positions. Then, a hip spica cast was applied at 20° of flexion, 30° of abduction and 10° of internal rotation for 6 weeks. After the removal of the cast, an abduction brace was used at all times for 6 weeks. At the final follow-up, clinical and radiological evaluations were performed using the modified McKay 7 and Severin 8 classification systems. The position of the femoral capital ossification centre on preoperative radiographs was assessed using the method of Tönnis 9. Avascular necrosis was classified on the basis of the criteria of Kalamchi and MacEwen 10. To evaluate the development of the acetabulum and the proximal femur, we measured the acetabular angle, the centre–edge angle and the neck–shaft angle of the femur on radiographs taken preoperatively, at the end of the first, third and fifth years, and at the final follow-up. The film focal length was 110 cm in all cases. All preoperative and postoperative radiographs were evaluated by the same author (M.E.B.) using the same goniometer.

angle and the neck–shaft angle of the femur on radiographs taken preoperatively, at the end of the first, third and fifth years, and at the final follow-up. The film focal length was 110 cm in all cases. All preoperative and postoperative radiographs were evaluated by the same author (M.E.B.) using the same goniometer. The statistical analyses were carried out using the IBM SPSS Statistics for Windows, version 22 (IBM Corp., Armonk, New York, USA). Percentage, rate, average and SD were used as descriptive statistics. The compliance of the quantitative data with normal distribution was evaluated using the Kolmogorov–Smirnov test. The parametric data were compared using the Student t-test. The χ2-test was used to assess the correlation between preoperative Tönnis hip scores and the presence of avascular necrosis at the latest follow-up. P value less than 0.05 was considered to be statistically significant. Results According to the Tönnis classification system, 20 hips were classified as grade II, nine were grade III and three were grade IV on the basis of the preoperative radiographs. There was no redislocation, loss of concentric reduction, infection or graft extrusion postoperatively. No patient required subsequent surgery (Fig. 1). Fig. 1 Anteroposterior pelvis views of a 17-month-old girl with right hip dysplasia. (a) Preoperative view. (b) After open reduction through the medial approach and PA. (c) Three years’ postoperative view. (d) Nine years later, with excellent clinical and radiological results. PA, Pemberton acetabuloplasty.

Results According to the Tönnis classification system, 20 hips were classified as grade II, nine were grade III and three were grade IV on the basis of the preoperative radiographs. There was no redislocation, loss of concentric reduction, infection or graft extrusion postoperatively. No patient required subsequent surgery (Fig. 1). Fig. 1 Anteroposterior pelvis views of a 17-month-old girl with right hip dysplasia. (a) Preoperative view. (b) After open reduction through the medial approach and PA. (c) Three years’ postoperative view. (d) Nine years later, with excellent clinical and radiological results. PA, Pemberton acetabuloplasty. At the final follow-up, clinical results were excellent in 30 hips (93.8%) and good in two (6.2%). Radiological assessment showed that 29 hips (90.6%) were rated as class I and three hips (9.4%) were class II (Fig. 2). The mean preoperative and postoperative radiological measurements are shown in Table 1. Fig. 2 Anteroposterior pelvis views of a 20-month-old girl with bilateral hip dislocation. (a) Preoperative view. (b) After bilateral open reduction through a medial approach and bilateral PA in a single session. (c) Five years after the operation. (d) 14 years later, with excellent clinical and radiological results. PA, Pemberton acetabuloplasty. Table 1 Mean (range) acetabular, centre–edge and neck–shaft angles before and after surgery

Fig. 2 Anteroposterior pelvis views of a 20-month-old girl with bilateral hip dislocation. (a) Preoperative view. (b) After bilateral open reduction through a medial approach and bilateral PA in a single session. (c) Five years after the operation. (d) 14 years later, with excellent clinical and radiological results. PA, Pemberton acetabuloplasty. Table 1 Mean (range) acetabular, centre–edge and neck–shaft angles before and after surgery Postoperatively, only three hips (9.3%) showed evidence of avascular necrosis. Two hips were classified as group I and one hip as group II according to the classification system of Kalamchi and MacEwen. All three hips showed very good radiological features at the final follow-up (two Severin class I and one Severin class II). At the latest follow-up, the centre–edge angles of the operated hips (mean 37.6°) were significantly higher than those of the unaffected hips (mean 29.5°) in unilateral operated patients (P<0.0001). There was no statistically significant difference between the acetabular angles of the operated hips (mean 10.8°) and those of the unaffected hips (mean 14.4°) in unilateral operated patients (P=0.427). Discussion The basic goal of the treatment in developmental dysplasia of the hip is to achieve a stable concentric reduction and to maintain the reduction during childhood and adolescence.

Postoperatively, only three hips (9.3%) showed evidence of avascular necrosis. Two hips were classified as group I and one hip as group II according to the classification system of Kalamchi and MacEwen. All three hips showed very good radiological features at the final follow-up (two Severin class I and one Severin class II). At the latest follow-up, the centre–edge angles of the operated hips (mean 37.6°) were significantly higher than those of the unaffected hips (mean 29.5°) in unilateral operated patients (P<0.0001). There was no statistically significant difference between the acetabular angles of the operated hips (mean 10.8°) and those of the unaffected hips (mean 14.4°) in unilateral operated patients (P=0.427). Discussion The basic goal of the treatment in developmental dysplasia of the hip is to achieve a stable concentric reduction and to maintain the reduction during childhood and adolescence. Open reduction through a medial approach enables direct access to the main obstacles to concentric reduction, such as tight iliopsoas tendon, inferomedial joint capsule, hypertrophied ligamentum teres and hypertrophied pulvinar. Thus, an intervention to correct for this obstacle can easily be performed with a medial approach. In our clinic, we have been using the Ferguson medial approach for 35 years and we believe that a medial approach is easier than an anterolateral approach for an experienced surgeon 11,12.

eres and hypertrophied pulvinar. Thus, an intervention to correct for this obstacle can easily be performed with a medial approach. In our clinic, we have been using the Ferguson medial approach for 35 years and we believe that a medial approach is easier than an anterolateral approach for an experienced surgeon 11,12. Avascular necrosis is one of the most feared complications of the treatment for developmental dysplasia of the hip. Reported rates of the incidence of avascular necrosis after medial open reduction vary from 0 to 67% 2,3,11–15. Extensive capsular dissection, T-shaped capsular incision and ligation of femoral circumflex vessels during the operation were considered responsible for increasing the incidence of avascular necrosis 11. However, some recent studies have reported that the risks of avascular necrosis are not significantly different between medial open reduction and anterolateral open reduction at early to mid-term follow-up periods 16,17. Avascular necrosis was acceptably low, with a 9.3% rate, in our study. We took great care to avoid damaging the medial circumflex vessels and we only performed a longitudinal inferomedial capsular incision. Moreover, no patient received previous failed treatment, the ossific nucleus was present before the operation in all hips and we practiced the PA without capsular dissection. We believe that all these precautions contributed towards the low avascular necrosis rates.

formed a longitudinal inferomedial capsular incision. Moreover, no patient received previous failed treatment, the ossific nucleus was present before the operation in all hips and we practiced the PA without capsular dissection. We believe that all these precautions contributed towards the low avascular necrosis rates. One criticism of the medial approach is that it may require secondary interventions for residual dysplasia. The rate of additional operations varies in the literature. Kiely et al. 18 reported that 22.4% of hips reduced using a Ferguson medial approach required further pelvic or femoral osteotomies over a mean 82-month follow-up period. Koizumi et al. 14 reported a 45.7% rate of additional operations after open reduction through a medial approach. Similarly, Sener et al. 19 have found the incidence of secondary procedures after medial open reduction of up to 70% in children older than 18 months of age. Salter and Dubos 20 recommended a pelvic osteotomy at the time of open reduction for correction of residual dysplasia because they believed that the potential for acetabular remodelling declines rapidly after 18 months of age 21. In contrast, Albinana et al. 22 reported that acetabular improvement may continue up to 6 years after reduction. Baki et al. 11 had previously reported satisfactory results of single-stage medial open reduction and Salter innominate osteotomy in our clinic. They reported no secondary interventions for residual dysplasia or redislocation over a mean 9.6-year follow-up period. Furthermore, it was shown that adding a pelvic osteotomy after the open reduction in the same session for the treatment of developmental dysplasia of the hip significantly reduced the incidence of subsequent surgery 23,24.

dary interventions for residual dysplasia or redislocation over a mean 9.6-year follow-up period. Furthermore, it was shown that adding a pelvic osteotomy after the open reduction in the same session for the treatment of developmental dysplasia of the hip significantly reduced the incidence of subsequent surgery 23,24. We combined the medial open reduction with PA in a single-stage procedure in patients who had an acetabular angle greater than 35° on the preoperative radiograph or instability after medial open reduction for preventing residual dysplasia. To our knowledge, this is the first report of a combination of Ferguson medial open reduction with PA in a single-stage operation.

single-stage procedure in patients who had an acetabular angle greater than 35° on the preoperative radiograph or instability after medial open reduction for preventing residual dysplasia. To our knowledge, this is the first report of a combination of Ferguson medial open reduction with PA in a single-stage operation. PA also has the advantage that patients with bilateral dislocations can be operated bilaterally in a single session, in contrast with the Salter osteotomy. A disadvantage of the Ferguson medial approach is the inability to perform a capsulorrhaphy; this can increase the risk of instability. In six hips, there was an instability after open reduction in our series. In all unstable hips, the acetabular angle was greater than 30°. According to the literature, the combination of instability and acetabular dysplasia may increase the requirement for additional corrective operations in this age group 13–15. Therefore, we preferred PA instead of capsuloplasty in these unstable hips to avoid subsequent operations. In our series, no patient required subsequent operation for subluxation, redislocation or residual dysplasia. We consider that PA provides sufficient stability to the hip joint and removes the disadvantage of the absence of capsulorrhaphy in the Ferguson medial approach.

se unstable hips to avoid subsequent operations. In our series, no patient required subsequent operation for subluxation, redislocation or residual dysplasia. We consider that PA provides sufficient stability to the hip joint and removes the disadvantage of the absence of capsulorrhaphy in the Ferguson medial approach. Femoral shortening osteotomies were usually performed if the surgeon judged the reduction to be under excessive tension 25. Furthermore, femoral shortening was recommended in older patients (>36 months of age) and high dislocations 26. We did not perform femoral shortening in any patient because our eldest patient was 24 months of age. Also, there was no abnormal tension in any hip joint after open reduction and PA. According to Mc Kay’s classification system, an excellent result is painless, stable hip with a negative Trendelenburg test and full range of motion. Clinically, 93.8% of patients were classified as showing an excellent result at the final follow-up in our series. Performing the PA without anterior open reduction enables a smaller anterolateral incision and less extensive anterior dissection (no anterior capsular dissection). We believe that these features may prevent the formation of fibrosis and adhesions around the hip joint. The disadvantage of this combined technique is the use of two separate incisions. We conclude that satisfactory clinical and radiological results can be achieved with a single-stage Ferguson medial open reduction and PA after the age of 15 months for the treatment of developmental dysplasia of the hip.

According to Mc Kay’s classification system, an excellent result is painless, stable hip with a negative Trendelenburg test and full range of motion. Clinically, 93.8% of patients were classified as showing an excellent result at the final follow-up in our series. Performing the PA without anterior open reduction enables a smaller anterolateral incision and less extensive anterior dissection (no anterior capsular dissection). We believe that these features may prevent the formation of fibrosis and adhesions around the hip joint. The disadvantage of this combined technique is the use of two separate incisions. We conclude that satisfactory clinical and radiological results can be achieved with a single-stage Ferguson medial open reduction and PA after the age of 15 months for the treatment of developmental dysplasia of the hip. Acknowledgements Conflicts of interest There are no conflicts of interest.

Introduction Unilateral congenital muscular torticollis, also known as wry neck, is a common condition caused by shortening and fibrosis of the sternocleidomastoid muscle (SCM). This condition is usually discovered during the first few weeks of life. It is characterized by the tilt of the infant’s head to one side, with difficulties turning to the opposite side. A nontender mass attached of the SCM at the side toward the head is tilting can also be noted. The bilateral contracture of the SCM is a very rare condition. We present a case of bilateral muscular torticollis first diagnosed and operated in a child aged 12 and a 25-year follow-up. Case report A 12-year-old boy was referred to the orthopedic department from a general-doctor practice. He was a second child of a nonconsanguineous marriage from full-term vaginal delivery in breech position. There was no family history of congenital deformities. He showed no other abnormalities. That particular patient had not been treated by any doctor before. All data considering his health status were provided by his parents, who reported that first symptoms of torticollis appeared in a preschool age.

n breech position. There was no family history of congenital deformities. He showed no other abnormalities. That particular patient had not been treated by any doctor before. All data considering his health status were provided by his parents, who reported that first symptoms of torticollis appeared in a preschool age. The physical examination of the child indicated asymmetry of the head and face and a smaller right cheek. The lines of the mouth and eyes were crossing on the right part of the face. The head presented an altered, compulsory position of bending to the right side with the chin rotated to the left by ∼20° (Figs 1–3). The observed aggravated tension of both SCM widened the shape of the neck and restricted its movements. The difference in shortening on both SCM was 2 cm. Range of motion (ROM) in the neck was limited for rotation to 30° and for lateral flexion to 20°. ROM on the left side was better and was 40° for rotation and 30° for lateral flexion. Infants of 2 months of age had a median muscle function score of 1, which makes it difficult for the baby to turn the head side to side. Examination of the spine showed compensatory deformation in the cervical and upper thoracic part, an increased thoracic kyphosis, and elevation of the right shoulder. Radiographs did not show any structural deformation of the vertebrae (Figs 4 and 5). As a treatment, the upper and lower tenotomies of the right SCM were performed and the child was immobilized in Schanz’s cervical orthosis. Three months after the operative procedure, the physical examination indicated an increasing contracture of the left SCM with time. The head again presented an altered, compulsory position of bending now to the left side and the chin rotation to the right. A similar operative procedure was applied to the left SCM. The child was placed in cervical orthosis in maximal extension, without any rotation or head lateral bending. The follow-up examinations during a period of 6 months showed good wound healing and a positive outcome. The child and the parents were satisfied with the result of the treatment.

e was applied to the left SCM. The child was placed in cervical orthosis in maximal extension, without any rotation or head lateral bending. The follow-up examinations during a period of 6 months showed good wound healing and a positive outcome. The child and the parents were satisfied with the result of the treatment. Fig. 1 Frontal view – age 12. Fig. 2 Back view – age 12. Fig. 3 Lateral view – age 12. Fig. 4 Anteroposterio view – radiography. Fig. 5 Lateral view – radiography. The patient was invited for re-examination 25 years after the initial treatment. The physical examination at the age of 37 years showed symmetrical rotation of the head and no restriction to flexion or to extension of the cervical spine (Figs 6 and 7). The ROM for rotation was to 90° on both sides. ROM for lateral flexion showed its highest range to about 60° both sides. The initial asymmetry of the face was hardly noticed. The patient is an agricultural physical worker. The previous treatment has no present effect on his well-being. Fig. 6 Frontal view – age 37. Fig. 7 Back view – age 37.

The patient was invited for re-examination 25 years after the initial treatment. The physical examination at the age of 37 years showed symmetrical rotation of the head and no restriction to flexion or to extension of the cervical spine (Figs 6 and 7). The ROM for rotation was to 90° on both sides. ROM for lateral flexion showed its highest range to about 60° both sides. The initial asymmetry of the face was hardly noticed. The patient is an agricultural physical worker. The previous treatment has no present effect on his well-being. Fig. 6 Frontal view – age 37. Fig. 7 Back view – age 37. Discussion Congenital muscular torticollis, or wry neck, is described as the most common form of painless torticollis. First descriptions were reported by Cheselden in 1749 1 and later by Anderson 1893 2. Unilateral SCM contracture causing torticollis is well known. However the etiology of CMT remains unknown. Possible causes include a lack of space in utero, resulting in local compartment syndrome or ischemia producing fibrotic SCM 3. As the child’s head is tilted toward the involved neck muscle and the chin rotated toward the contralateral shoulder, the diagnosis is made on physical examination at birth or shortly after 1,4. The finding of a hardened mass of SCM muscle on the involved side facilitates diagnostics. The unilateral contracture is almost universally present after infancy and may be replaced later with a fibrous contracted band 5.

ntralateral shoulder, the diagnosis is made on physical examination at birth or shortly after 1,4. The finding of a hardened mass of SCM muscle on the involved side facilitates diagnostics. The unilateral contracture is almost universally present after infancy and may be replaced later with a fibrous contracted band 5. Many authors agree that the nonsurgical treatment with stretching and massage yields an excellent result in more than 90% of patients 1,6–10. Others emphasize the fact that for such treatment to be very successful, it should be started within a few months after birth. If this or other treatments do not yield positive results, surgery can sometimes correct the problem. Conservative management is rarely successful in patients presented in childhood 7,11. Less than 10% of cases require a surgical approach if a significant restriction of motion (above 30°) or facial asymmetry is present at school age 4,10,12. Ferkel bipolar lengthening of SCM is a procedure of choice for those patients 13. We are of the opinion that in the patients seen in childhood, surgical intervention should be considered the treatment of choice to avoid further irreversible changes.

tion (above 30°) or facial asymmetry is present at school age 4,10,12. Ferkel bipolar lengthening of SCM is a procedure of choice for those patients 13. We are of the opinion that in the patients seen in childhood, surgical intervention should be considered the treatment of choice to avoid further irreversible changes. Bilateral congenital muscular torticollis is a very rare form of congenital muscular torticollis. The medical literature is scarce on cases of bilateral torticollis, with just three cases described in Hungary in 1993 14, in Singapore in 2009 15, and in China in 2016 16. A birth injury or a defective embryogenesis could likely be a cause of unilateral condition. Other theories of the origin of bilateral torticollis include fibrosis of the SCM muscle, resulting from venous occlusion because of intrauterine persistent position. Finally, we should take into consideration a genetic, growth retardation, infectious myositis, or mix of a variety of factors caused. Torticollis is a well-known deformity; however, bilateral malformation might present a differential diagnostic problem to orthopedic surgeons as well as ophthalmologists and neurologists. Acknowledgements The authors have contributed substantially to the submitted work, have made a significant contribution to the findings and methods in the paper and have reviewed and agree with the submission of the manuscript for review. All authors have read and approved the final draft. Conflicts of interest There are no conflicts of interest.

Introduction In-toeing gait is a common feature of cerebral palsy (CP) children that causes poor cosmesis and function 1–3. In-toeing is the result of abnormal muscle tone, internal torsion of the femur and tibia, and various foot deformities 4,5. Torsional deformity of the femur or the tibia causes lever-arm disease and reduces the efficiency of muscles during walking 6. Among these, medial femoral torsion (MFT) is the most common pathology of in-toeing gait in CP children.

muscle tone, internal torsion of the femur and tibia, and various foot deformities 4,5. Torsional deformity of the femur or the tibia causes lever-arm disease and reduces the efficiency of muscles during walking 6. Among these, medial femoral torsion (MFT) is the most common pathology of in-toeing gait in CP children. Surgical correction of MFT is recommended in CP children to resolve the inefficient lever-arm of torsional deformity 1. Femoral derotational osteotomy (FDO) could correct MFT in CP children and directly improve the foot progression angle (FPA) and hip rotation angle in the transverse plane. In addition, it is indicated when in-toeing is principally because of MFT. Many authors have reported that various gait parameters improve or change after FDO to varying degrees 1,3,7,8. There are, however, few papers on correlations between the amount of derotation by FDO and the degree of changes in gait parameters after FDO 9. FDO can be performed proximally and distally. The proximal FDO is frequently performed to decrease the increased anteversion and at the same time to correct other hip problems, such as valgus angulation of femoral neck, subluxation of hip joint, and others. There may be some bias when analyzing the simple correlation between the derotation angle by proximal FDO and postoperative kinematic changes in CP patients with hip problems 1. The aim of this study was to analyze how much the kinematic data and Staheli’s rotational profiles changed after a certain degree of distal FDO in CP patients.

bias when analyzing the simple correlation between the derotation angle by proximal FDO and postoperative kinematic changes in CP patients with hip problems 1. The aim of this study was to analyze how much the kinematic data and Staheli’s rotational profiles changed after a certain degree of distal FDO in CP patients. Patients and methods Study population This retrospective analysis included CP patients younger than 20 years of age who underwent distal FDO to correct MFT between 2010 and 2013. The average follow-up period was 3.8 years (2–7.5 years). The study group included 19 spastic CP patients (28 lower limbs) with the Gross Motor Function Classification System level I or II. All the patients underwent a derotational osteotomy in the distal femur. MFT and gait of the patients were evaluated with Staheli’s rotational profile and gait kinematics. A preoperative evaluation was performed 1 day before surgery and a postoperative evaluation was performed 1 year after surgery. Exclusion criteria were as follows: subluxated hip, previous selective dorsal rhizotomy, and injections of botulinum toxin A in the previous 6 months. Also, CP patients who had been treated for long bone fractures or for torsional deformities in segments (tibia or foot) other than the femur were excluded.

fter surgery. Exclusion criteria were as follows: subluxated hip, previous selective dorsal rhizotomy, and injections of botulinum toxin A in the previous 6 months. Also, CP patients who had been treated for long bone fractures or for torsional deformities in segments (tibia or foot) other than the femur were excluded. The study group was divided into the bilateral group (BG) (nine cases, CP spastic diplegia patients) and the unilateral group (UG) (10 cases, usually CP spastic hemiplegia patients) for analysis. At the time of surgery, the mean age of the study group was 13.2 years (range: 4–20, the BG: 12.5 years, and UG: 13.9 years). There were 10 male and nine female patients in the study group. There were four male patients in the BG and six patients in the UG (Table 1). Table 1 Patients’ profile in the study group Derotational osteotomy All surgeries were performed by a single senior surgeon. Surgical treatment of each patient was planned according to the principles of single-event multilevel surgery by Gage or SMILE by Sussman, and various operations in different segments were performed simultaneously, including FDO (Table 1). Distal FDO was indicated for patients with symptoms and increased MFT. Patients’ symptoms included tripping or a patient susceptible to fall down during walking. FDO was indicated with limited external rotation of the hip less than 15° on physical examination and hip internal rotation and internal FPA with a difference of more than 10° from the norm on kinematic data.

s and increased MFT. Patients’ symptoms included tripping or a patient susceptible to fall down during walking. FDO was indicated with limited external rotation of the hip less than 15° on physical examination and hip internal rotation and internal FPA with a difference of more than 10° from the norm on kinematic data. Preoperatively, we decided on the amount of derotation using various data such as Staheli’s rotational profile, computed tomography (CT) torsional study, and kinematics of gait analysis. A CT torsional study was carried out as a routine preoperative evaluation together with the gait analysis. However, the CT torsional study was not carried out postoperatively, unlike the gait analysis, and was excluded from statistical analysis. The mean amount of derotation was 24.6° by the FDO (a mean of 25.0° for the UG and a mean of 24.4° for the BG).

as a routine preoperative evaluation together with the gait analysis. However, the CT torsional study was not carried out postoperatively, unlike the gait analysis, and was excluded from statistical analysis. The mean amount of derotation was 24.6° by the FDO (a mean of 25.0° for the UG and a mean of 24.4° for the BG). Derotational osteotomy was performed at the distal femur in the supine position. The distal femur was approached on the lateral side. Two guide wires (Kirschner wire, 1.6 mm) were inserted tilted as much as the predetermined correction angle proximal and distal to the osteotomy site. After the osteotomy, the distal segment was externally rotated so that the two guide wires were positioned parallel to one another. Then, the osteotomy site was internally fixated using an AO locking compression plate (LCP). Narrow or broad AO LCP was chosen according to the patient’s femur size. A cylinder splint was applied for 2 weeks postoperatively. Knee range-of-motion exercise was started 1 week after surgery. Partial weight bearing was allowed 2 weeks after surgery. Evaluation of deformity The effects of FDO were assessed by comparing preoperative Staheli’s rotational profile and gait kinematics with postoperative values. Staheli’s rotational profile was checked at the time of gait analysis. Among the five parameters of Staheli’s rotational profile, the degree of internal and external rotation of the hip joint, which reflects the femoral torsion, was used for evaluation 10,11. Hip rotation was measured in the prone position using a goniometer.

’s rotational profile was checked at the time of gait analysis. Among the five parameters of Staheli’s rotational profile, the degree of internal and external rotation of the hip joint, which reflects the femoral torsion, was used for evaluation 10,11. Hip rotation was measured in the prone position using a goniometer. Gait analysis was carried out preoperatively and postoperatively. A preoperative gait analysis was carried out 1 day before the surgery and a postoperative gait analysis was carried out in patients after obtaining their informed consent when they were readmitted for the removal of internally fixated implants. The plate and screws were removed 1 year after surgery in most cases. For gait analysis, passive reflective markers were attached using the modified Helen-Hayes method. Optical tracking of reflective markers was performed using the Eagle camera system (Eagle system; Motion Analysis, Santa Rosa, California, USA) at a capture frequency of 120 frames per second. EvaRT version 4.2 (Motion Analysis) and Orthotrak (Motion Analysis) softwares were used for data processing to obtain kinematic data from the three-dimensional gait analysis.

ormed using the Eagle camera system (Eagle system; Motion Analysis, Santa Rosa, California, USA) at a capture frequency of 120 frames per second. EvaRT version 4.2 (Motion Analysis) and Orthotrak (Motion Analysis) softwares were used for data processing to obtain kinematic data from the three-dimensional gait analysis. Statistical analysis We analyzed the amount of changes in various parameters after a certain amount of derotation by FDO (mean of 24.6°; 25.0° for the UG and 24.4° for the BG). The paired t-test was used to analyze changes in rotational profiles and kinematic data after FDO. Linear regression was used to evaluate the amount of change after derotational osteotomy. Statistical analysis of the correlation was carried out using the Pearson correlation test, with a 95% confidence interval (P<0.05) for statistical significance. The SPSS (version 18.0; SPSS Inc., Chicago, Illinois, USA) program was used. Results Staheli’s rotational profile The mean internal hip rotation was 71.4±6.9° before surgery and 48.6±10.7° after surgery in the UG, meaning that a decrease of 22.8° occurred after surgical derotation of 25.0° (P<0.05). A decrease of internal rotation in the BG was 22.9° (P<0.05), with 63.8±15.8° before surgery and 40.9±9.2° after surgical derotation of 24.4°. The changes in internal rotation in both groups (22.8°, 22.9°) were almost proportional to the amount of surgical derotation (24.6°) (Table 2). Table 2 Postoperative changes in the rotational profile and kinematics after femoral derotational osteotomy

Results Staheli’s rotational profile The mean internal hip rotation was 71.4±6.9° before surgery and 48.6±10.7° after surgery in the UG, meaning that a decrease of 22.8° occurred after surgical derotation of 25.0° (P<0.05). A decrease of internal rotation in the BG was 22.9° (P<0.05), with 63.8±15.8° before surgery and 40.9±9.2° after surgical derotation of 24.4°. The changes in internal rotation in both groups (22.8°, 22.9°) were almost proportional to the amount of surgical derotation (24.6°) (Table 2). Table 2 Postoperative changes in the rotational profile and kinematics after femoral derotational osteotomy The mean external hip rotation in the UG was −14.3±15.9° before surgery and −41.4±10.3° after surgery (a minus sign denotes ‘external’ in this study), showing an increase of 27.1° (P<0.05), which is the change almost equal to the amount of surgical derotation (25.0°). In the BG, the mean external hip rotation increased by about 16.6° (P<0.05), with −32.5±14.5° before surgery and −49.1±16.4° after surgery, which is almost two-thirds of surgical derotation (24.4°). Gait kinematics Foot progression angle and hip rotation The mean FPA (a minus sign indicates ‘external’ for the kinematic data) was −2.1° during gait in the UG before surgery and −13.8° after surgical derotation of 25.0° (P<0.05). In the BG, FPA was 3.1° preoperatively and −9.5° postoperatively (P<0.05); FPA was 12.6° more outward after FDO. In both groups, FPA showed changes in parameters equal to 50% of surgical derotation.

he kinematic data) was −2.1° during gait in the UG before surgery and −13.8° after surgical derotation of 25.0° (P<0.05). In the BG, FPA was 3.1° preoperatively and −9.5° postoperatively (P<0.05); FPA was 12.6° more outward after FDO. In both groups, FPA showed changes in parameters equal to 50% of surgical derotation. Hip rotation during gait is a parameter directly affected by the FDO. In the UG, the mean hip rotation during gait was 15.9° before surgery and 1.1° after surgery, which meant that the hip joint was rotated more externally by 14.8° after surgical derotation of 25.0° (P<0.05). The BG showed a hip rotation of 7.4° before surgery and an average of 0.7° after surgery, meaning that the hip joint rotated externally by 6.7° after derotation of the femur by 24.4° (P<0.05). Foot and knee rotation FDO is a surgical procedure on the femur and basically it does not affect the transverse kinematics of other segments, such as foot rotation or knee rotation. In the UG, the average foot rotation (a minus sign denotes ‘external’ for the kinematic data) was −4.1° preoperatively and −5.4° postoperatively. In the BG, it was 2.0° before surgery and −3.0° after surgery. In both groups, foot rotation was not affected by the FDO (P>0.05). Knee rotation mostly reflects tibial torsion, and in the UG, the mean knee rotation was −12.7° before surgery and −10.1° after surgery (P>0.05), which meant that it was not influenced by the FDO. However, in the BG, it was −4.7° before surgery and −14.3° after surgery, which meant that knee rotation was changed by 9.6° after FDO (P<0.05).

reflects tibial torsion, and in the UG, the mean knee rotation was −12.7° before surgery and −10.1° after surgery (P>0.05), which meant that it was not influenced by the FDO. However, in the BG, it was −4.7° before surgery and −14.3° after surgery, which meant that knee rotation was changed by 9.6° after FDO (P<0.05). Pelvic rotation In the UG, the mean pelvic rotation was 0.4° preoperatively and 0.6° postoperatively (P>0.05). In the BG, the mean pelvic rotation was 0.5° preoperatively and 0.2° postoperatively (P>0.05). Pelvic rotation was not changed after surgery in both groups. However, in the patients with compensatory pelvic rotation of more than 5° preoperatively, pelvic rotation showed a change of about 5.4±4.8° in the UG and 6.6±1.5° in the BG after surgery (P<0.05). There was also a trend indicating that the younger the patient, the more the pelvic rotation changed (P=0.069). The total range of pelvic rotation during gait was not meaningfully changed (16.0°–15.7°) in the UG (P>0.05). However, it showed a significant difference in the BG postoperatively (21.2°–15.4°; P<0.05). Complications There were no complications, such as nonunion or operative wound infection, in any of the patients.

However, in the patients with compensatory pelvic rotation of more than 5° preoperatively, pelvic rotation showed a change of about 5.4±4.8° in the UG and 6.6±1.5° in the BG after surgery (P<0.05). There was also a trend indicating that the younger the patient, the more the pelvic rotation changed (P=0.069). The total range of pelvic rotation during gait was not meaningfully changed (16.0°–15.7°) in the UG (P>0.05). However, it showed a significant difference in the BG postoperatively (21.2°–15.4°; P<0.05). Complications There were no complications, such as nonunion or operative wound infection, in any of the patients. Discussion Classifications, such as the classification of Winters and colleagues 12,13, of the pathologic gait in CP children are mostly focused on the sagittal plane. The most common pathologic gait of CP children in transverse plane is in-toeing gait caused by MFT. MFT causes lever arm dysfunction which reduces muscle efficiency by reducing the coronal plane hip abductor moment arm. And it also causes inappropriate foot clearance and tripping. 6,14.

y focused on the sagittal plane. The most common pathologic gait of CP children in transverse plane is in-toeing gait caused by MFT. MFT causes lever arm dysfunction which reduces muscle efficiency by reducing the coronal plane hip abductor moment arm. And it also causes inappropriate foot clearance and tripping. 6,14. It is well known from previous studies that MFT can be corrected with FDO. It is, however, not clear to what degree transverse gait parameters, including hip rotation, FPA, pelvic rotation, and others, are changed after surgery. In this study, we attempted to answer these questions. FDO was performed at the distal femur and internally fixated with AO LCP. The advantage of LCP was that derotation of the distal segment could be performed accurately as planned and early rehabilitation could be started without immobilization in the long leg cast. Parameters that are directly affected by FDO are internal and external hip rotation in Staheli’s rotational profile and FPA and hip rotation in the gait analysis. Pirpiris et al. 1 reported that internal and external hip rotation in the prone position improved as much as the amount of FDO in CP patients, which is in agreement with this study. Chung et al. 15 showed that FPA improved by two-thirds of the amount of FDO after bilateral FDO in diplegic CP. In this study, FPA improved about half of the FDO in both groups. FPA improved less than the amount of FDO 15.

tion improved as much as the amount of FDO in CP patients, which is in agreement with this study. Chung et al. 15 showed that FPA improved by two-thirds of the amount of FDO after bilateral FDO in diplegic CP. In this study, FPA improved about half of the FDO in both groups. FPA improved less than the amount of FDO 15. The change in FPA is roughly the sum of rotation changes in all the segments of lower extremity. In the UG, improvement in FPA occurred entirely at the femur (hip rotation), which was equal to half of FDO. In the BG, superficially, it appeared as if half of the FPA improvement (6.7°) took place at the femur (hip rotation) and another half of the improvement (9.6°) took place at the tibia (knee rotation) according to the kinematic data. Knee rotation was not changed postoperatively in the UG, as expected. In the BG, knee rotation changed by as much as one-fourth of FDO, even though theoretically it should not be affected by the FDO. Transverse kinematic data of the BG were confusing as the improvement in FPA took place at the femur and tibia. This might be because of measuring error that is prone to occur in the transverse plane compared with other planes. There are still debates on whether compensatory pelvic rotation is corrected after FDO. Some authors, including Chung et al. 15 and Christopher et al. 16, reported that pelvic rotation in spastic diplegia was not changed even after normalization of femoral torsion by FDO.

He recommended a slight undercorrection during FDO for hemiplegias because of the above-mentioned reason. Meanwhile, Christopher et al. 16 reported that hemiplegic patients showed a significant improvement in pelvic rotation after FDO. In this study, the overall pelvic rotation was not changed after FDO in both groups. As the study group included many patients without compensatory pelvic rotation, we reanalyzed changes in pelvic rotation only in patients with preoperative pelvic rotation of more than 5°. In these patients, preoperatively externally rotated pelvis showed normalization of rotation after FDO in both the UG and the BG (P<0.05). In addition, the degree of change tended to be larger in younger patients. These facts suggest that compensatory pelvic rotation may be corrected after FDO if the deformity is not a fixed one as in young children. The range of pelvic rotation is a parameter that reflects the energy efficiency of ambulation. The fact that range of rotation has increased means that additional energy was used to transfer the center of gravity to achieve an adequate stride length. There was no major change in the range of rotation in the hemiplegia group (0.3° decrease, P=0.87), but there was a statistically significant decrease in the diplegia group (5.8° decrease, P<0.05, Figs 1–3). Therefore, a decrease in the range of pelvic motion after FDO might be an indication that FDO may improve energy efficiency during gait. Fig. 1 The distal segment was accurately derotated using two guide pins and was securely fixated using AO locking compression plate.

The range of pelvic rotation is a parameter that reflects the energy efficiency of ambulation. The fact that range of rotation has increased means that additional energy was used to transfer the center of gravity to achieve an adequate stride length. There was no major change in the range of rotation in the hemiplegia group (0.3° decrease, P=0.87), but there was a statistically significant decrease in the diplegia group (5.8° decrease, P<0.05, Figs 1–3). Therefore, a decrease in the range of pelvic motion after FDO might be an indication that FDO may improve energy efficiency during gait. Fig. 1 The distal segment was accurately derotated using two guide pins and was securely fixated using AO locking compression plate. Fig. 2 Gait parameters in the transverse plane improved to varying degrees after 24.6° of surgical derotation. Fig. 3 Ranges of pelvic rotation decreased after femoral derotational osteotomy in the bilateral group. There are some limitations in this study. First, we did not analyze the effect of soft tissue procedures, such as psoas or hamstring release, and foot procedures, such as split tibialis anterior tendon transfer, heel cord lengthening, or calcaneal lengthening. However, they may mainly influence motion in the sagittal plane and seem to have a slight effect in the transverse plane. Second, two young patients (4 and 5 years old) were included in this study. This may have affected the results. Third, our results were analyzed 1 year after operation. Hence, we do not have long-term follow-up results. Further research is needed.

sagittal plane and seem to have a slight effect in the transverse plane. Second, two young patients (4 and 5 years old) were included in this study. This may have affected the results. Third, our results were analyzed 1 year after operation. Hence, we do not have long-term follow-up results. Further research is needed. Conclusions and clinical significance In-toeing gait because of MFT could be corrected with FDO. FPA and hip rotation on the kinematics were corrected and the amount of correction was about a half of the surgical derotation angle. The degree of compensatory pelvic rotation should be considered to plan the derotation and decide the correction angle. Patients with pelvic rotation larger than 5° before the surgery showed the normalization of pelvic rotation after FDO. This phenomenon was obvious in young patients. Pelvic rotation in aged patients, whose pelvic rotation is considered to be already fixed, did not significantly change even after a surgery. Future research on FDO impact on clinical practice and the patient`s quality of life is needed. Acknowledgements Conflicts of interest There are no conflicts of interest.

Introduction Femoral neck fracture in children is a very rare event and comprises less than 1% of all pediatric fractures 1. The reason for its rarity is that the femoral neck in children is covered with a durable periosteum, which makes it strong and thick 2. Therefore, a femoral neck fracture in children signifies high-energy damage and its treatment requires caution 2. Not only is this type of fracture rare, but it also has a high rate of complications 2,3. Varus angulation, delayed union, and nonunion are related to surgical treatments, whereas avascular necrosis (AVN), premature closure of the epiphyseal plate, and lower limb paralysis are related to nonsurgical treatments. Out of these complications, which affects the final outcome the most and also occurs most frequently, is AVN of the femoral head 4–6. If AVN of the femoral head occurs, the subchondral bone weakens and fracture occurs. Subchondral fracture of the femoral head (crescent sign) causes pain and articular rigidity. Also, it pathologically causes flattening of the femoral head, and thus deformation of the femoral head and arthritis may develop 7,8.

ad 4–6. If AVN of the femoral head occurs, the subchondral bone weakens and fracture occurs. Subchondral fracture of the femoral head (crescent sign) causes pain and articular rigidity. Also, it pathologically causes flattening of the femoral head, and thus deformation of the femoral head and arthritis may develop 7,8. If AVN occurs after femoral neck fracture, appropriate treatment is needed according to the age of the patient or the degree of necrosis to minimize deformity of the femoral head. Primarily, containment treatment (the use of a brace or varus osteotomy) or valgus intertrochanteric osteotomy can be performed 9. Children with AVN of the femoral head might show abnormal painful hinge movement of the hip, which is because of impingement of the protruded anterolateral osteocartilaginous hump of the deformed femoral head against the acetabular rim. Valgus femoral osteotomy could be indicated in this condition as it moves the ‘hump’ on the femoral head further away from the acetabular margin, thus alleviating the hinge abduction, and effectively preventing the progression of femoral head deformity and arthritis 10. In children who are older than 10 years, if widespread AVN of the femoral head occurs, distraction arthroplasty rather than osteotomy may be performed, but the treatment outcome is not very favorable in most cases, and artificial hip replacement or articular fusion may need to be considered 11–13. However, taking into account the fact that the patient was not yet an adult, even these options were not feasible.

ction arthroplasty rather than osteotomy may be performed, but the treatment outcome is not very favorable in most cases, and artificial hip replacement or articular fusion may need to be considered 11–13. However, taking into account the fact that the patient was not yet an adult, even these options were not feasible. We report a case of femoral head AVN occurring after a pediatric femoral neck fracture that was treated with femoral head wedge resection and showed good results. Case presentation A 9-year-old girl presented to the emergency room complaining of right hip pain that occurred after a pedestrian car accident. She was hemodynamically stable and no other accompanying damage was observed. The hip radiography and computed tomography scans taken in the emergency room showed a displaced Delbet type 2 fracture of the right femoral neck (Fig. 1). Emergency surgery was performed under anesthesia, and closed reduction and internal fixation using a cannulated screw were performed (Fig. 2a). A cast was placed for 6 weeks, and bone union was achieved 3 months after the surgery (Fig. 2b). Fig. 1 (a) Bilateral pelvic anteroposterior radiograph taken in the emergency room. A right femoral neck fracture is observed. (b) The three-dimensional computed tomography shows translocation and angular deformity of the right femoral neck fracture.

Case presentation A 9-year-old girl presented to the emergency room complaining of right hip pain that occurred after a pedestrian car accident. She was hemodynamically stable and no other accompanying damage was observed. The hip radiography and computed tomography scans taken in the emergency room showed a displaced Delbet type 2 fracture of the right femoral neck (Fig. 1). Emergency surgery was performed under anesthesia, and closed reduction and internal fixation using a cannulated screw were performed (Fig. 2a). A cast was placed for 6 weeks, and bone union was achieved 3 months after the surgery (Fig. 2b). Fig. 1 (a) Bilateral pelvic anteroposterior radiograph taken in the emergency room. A right femoral neck fracture is observed. (b) The three-dimensional computed tomography shows translocation and angular deformity of the right femoral neck fracture. Fig. 2 (a) Hip radiograph taken immediately following surgery. Internal fixation with two cannulated screws was performed without damaging the epiphyseal plate. (b) The hip radiograph taken 3 months after surgery shows bone union of the fracture site. (c) Eight months after surgery, femoral head depression with avascular necrosis was observed. (d) One year after surgery, femoral head avascular necrosis had progressed and the metal screw was removed.

aging the epiphyseal plate. (b) The hip radiograph taken 3 months after surgery shows bone union of the fracture site. (c) Eight months after surgery, femoral head depression with avascular necrosis was observed. (d) One year after surgery, femoral head avascular necrosis had progressed and the metal screw was removed. Eight months after the surgery, the patient complained of right hip pain that worsened with activity, and the radiograph showed femoral head depression accompanied by AVN (Fig. 2c). The metal screw was removed 1 year after the surgery, but the AVN of the femoral head continued to progress (Fig. 2d). Pain of the hip joint continued to worsen; therefore, a second surgery was scheduled and an arthrogram was performed under anesthesia. The hip joint showed hinged abduction, which leads to the best congruency in the adduction–internal rotation position. Valgus-derotation-extension intertrochanteric osteotomy, which can best correct the hip congruency, was performed (Fig. 3). After the surgery, the pain was temporarily alleviated, but deformation of the joint continued to progress, and 2.5 years after injury, she complained of pain after several minutes of walking. She showed an antalgic gait pattern. On physical examination, hip abduction was 20° and hip internal rotation in the prone position was 15°. The follow-up radiograph showed more progression of the femoral head depression, because of which the sphericity of the femoral head was lost and a bump had formed on the anterolateral portion of the head (Fig. 4).

. On physical examination, hip abduction was 20° and hip internal rotation in the prone position was 15°. The follow-up radiograph showed more progression of the femoral head depression, because of which the sphericity of the femoral head was lost and a bump had formed on the anterolateral portion of the head (Fig. 4). Fig. 3 (a) Bilateral pelvic anteroposterior radiograph taken after valgization-derotation-extension osteotomy. (b) Translateral view of the right hip taken immediately after the second surgery. Fig. 4 Right pelvic anteroposterior radiograph taken 1 year after the second surgery (a) and the translateral view (b). A third surgery was performed to recover the sphericity of the femoral head and to increase the range of motion. Surgery was performed in the semilateral position with the sandbag under the ipsilateral hip. We used an anterolateral approach (Watson–Jones) for wedge resection. A T-shaped incision was made for capsulotomy. While abducting-adducting and rotating the joint, we first assessed the pathologic shape of the femoral head. There was a longitudinal fissure along the lateral articular surface of the femoral head, denuded of normal joint cartilage, which was created by impingement of the large femoral head against the lateral margin of the acetabulum.

ucting and rotating the joint, we first assessed the pathologic shape of the femoral head. There was a longitudinal fissure along the lateral articular surface of the femoral head, denuded of normal joint cartilage, which was created by impingement of the large femoral head against the lateral margin of the acetabulum. Wedge of the bone was resected including the area of longitudinal fissure (Fig. 5a and b). Then, the base of the lateral bump was greenstick fractured to be repositioned in the previously fissured area, which reduced the size of the femoral head. We checked the adequacy of wedge excision for the hip joint to move without impingement between bump on the femoral head and the acetabular edge or hinged abduction that the femoral head is hinged to the acetabulum and the medial joint space widening occurs. Fig. 5 (a) Picture taken during surgery of the right hip joint. The black arrow is the femoral head and the white arrow is the resected bone fragment. (b) The white arrow is the direction of movement of the lateral fragment. (c) Bilateral pelvic anteroposterior radiograph taken after femoral head wedge resection. (d) Frog leg view of the right hip taken after femoral head wedge resection. The empty space caused by the resection was filled so that sphericity of the femoral head was recovered, and internal fixation was performed using a headless screw. The radiographs taken after the surgery showed recovery of the sphericity of the femoral head (Fig. 5c and d).

Fig. 5 (a) Picture taken during surgery of the right hip joint. The black arrow is the femoral head and the white arrow is the resected bone fragment. (b) The white arrow is the direction of movement of the lateral fragment. (c) Bilateral pelvic anteroposterior radiograph taken after femoral head wedge resection. (d) Frog leg view of the right hip taken after femoral head wedge resection. The empty space caused by the resection was filled so that sphericity of the femoral head was recovered, and internal fixation was performed using a headless screw. The radiographs taken after the surgery showed recovery of the sphericity of the femoral head (Fig. 5c and d). After the wedge resection, pain and limping decreased rapidly, and the patient was satisfied with the surgical results. Hip range of motion increased and was painless postoperatively. Hip abduction and internal rotation improved to 40°. She could sit crossed-legged on the floor comfortably. The patient does not complain of pain or limping during the 3 years follow-up after the last operation. The metal screw was removed 2 years after resection, and the patient did not complain of any motor disturbance or pain. Also, the follow-up radiograph showed appropriate and satisfactory congruency and containment (Fig. 6). Fig. 6 (a) Bilateral pelvic anteroposterior radiograph taken 1.5 years after femoral head wedge resection and removal of the metal screws. (b) Translateral view of the right hip after metal screw removal.

The metal screw was removed 2 years after resection, and the patient did not complain of any motor disturbance or pain. Also, the follow-up radiograph showed appropriate and satisfactory congruency and containment (Fig. 6). Fig. 6 (a) Bilateral pelvic anteroposterior radiograph taken 1.5 years after femoral head wedge resection and removal of the metal screws. (b) Translateral view of the right hip after metal screw removal. Discussion Reports of femoral head AVN vary, at 14–47% 2. The risk factors of femoral head AVN after a femoral neck fracture include age, degree of fracture displacement, fracture severity, time to surgery after injury, and method of fixation 4,14. Of these, the most important factor is the severity of vascular damage at the time of injury. In adults, the intraosseous blood vessels may provide blood supply to the femoral head, but in children, the vessels cannot cross the open physis; therefore, blood supply is easily cut off if a hip fracture occurs 2.

ion 4,14. Of these, the most important factor is the severity of vascular damage at the time of injury. In adults, the intraosseous blood vessels may provide blood supply to the femoral head, but in children, the vessels cannot cross the open physis; therefore, blood supply is easily cut off if a hip fracture occurs 2. According to previous reports, the occurrence of AVN after femoral neck fracture is an important factor that affects prognosis, and the outcome is usually unsatisfactory in patients who have the condition 2,15. However, unlike in adults, the number of treatment options is limited in children. Nonsurgical methods cannot stop progression. A hip replacement or articular fusion in patients with an open physeal plate can later cause problems in development or in outer appearance. Therefore, there are about two surgical options that can be performed. The first is a valgus intertrochanteric osteotomy 9. This involves shifting the weight-bearing surface to another area that is not affected by AVN, and can be attempted if the extent of necrosis is not too wide. However, as in this case, if widespread necrosis occurs, there is a limit to the normal surface and collapse cannot be prevented. In this case, despite the fact that valgus osteotomy was performed, necrosis led to a collapse and the formation of a bump, which restricted the range of motion and aggravated pain and required additional treatment. The second option is a trap-door operation (subchondral bone grafting for segmental collapse) 16. The limitations of this method are that if deformity because of collapse is severe, the surgical method may become more difficult to perform, and in areas that have poor circulation, the bone graft may not heal properly.

. The second option is a trap-door operation (subchondral bone grafting for segmental collapse) 16. The limitations of this method are that if deformity because of collapse is severe, the surgical method may become more difficult to perform, and in areas that have poor circulation, the bone graft may not heal properly. The femoral head wedge resection performed in this case, unlike other surgical methods, has the advantage that it can be performed even in cases with widespread necrosis, taking into account the degree of femoral head deformity. The necrotic depressed area collapsed area of the femoral head due to osteonecrosis can be directly removed and an additional bone graft can be performed in the remaining area. Also, the remaining undamaged articular surface can be used so that the progression of degenerative arthritis can be delayed, and other surgical methods such as intertrochanteric osteotomy can be performed concomitantly. Previous methods such as cheilectomy lead to exposure of the subchondral bone, not the articular cartilage after resection of the bump; therefore, it can be considered a type of salvage operation, but in the femoral head wedge resection (wedge resection-greenstick fracture-fixation) performed in this case, the femoral head remains covered by the articular cartilage, and thus is expected to have a better long-term prognosis.

after resection of the bump; therefore, it can be considered a type of salvage operation, but in the femoral head wedge resection (wedge resection-greenstick fracture-fixation) performed in this case, the femoral head remains covered by the articular cartilage, and thus is expected to have a better long-term prognosis. Rarely, femoral head osteotomy has been reported in past studies. Siebenrock et al. 17 performed head reduction osteotomy in patients with Legg–Calve–Perthes disease. They used the posterior approach and trochanteric osteotomy. A periosteal flap was made to preserve the relevant branch of the medial circumflex artery and osteotomy was performed in the necrotic depressed area. Also, complete osteotomy was performed on the lateral fragment. This altered the anatomy of the femoral neck. In a study by Burian et al. 18, the anterolateral approach was used and periosteum on the anterior aspect of the neck was incised. However, the wedge resection in this study was different from previous studies. We used the familiar anterolateral approach and the femoral head was not dislocated for the procedure. The wedge osteotomy was performed in the bump area and the lateral fragment was repositioned with a greenstick fracture. This had the advantage that bony blood flow through the neck was not completely blocked and postoperative immobilization was minimal. Also, there was no other damage that could affect vascularity after capsulotomy.

tomy was performed in the bump area and the lateral fragment was repositioned with a greenstick fracture. This had the advantage that bony blood flow through the neck was not completely blocked and postoperative immobilization was minimal. Also, there was no other damage that could affect vascularity after capsulotomy. Many different types of surgery are being attempted in the treatment of femoral head AVN occurring after a pediatric femoral neck fracture, but there is still no consensus on the optimal treatment method 9. However, the femoral head wedge resection performed in this case is the only surgical procedure that can overcome the disadvantages of other surgeries the disadvantages of other surgeries. This method can be performed concomitantly or additional to other surgeries, and can help achieve appropriate congruency and containment of the hip joint. Acknowledgements Conflicts of interest There are no conflicts of interest.

Introduction Calcaneal fractures are probably the most common fractures in the tarsal bones, accounting for 60% of all tarsal fractures 1. Some calcaneal fractures are minor injuries, but many are severe, high-energy fractures. These more serious injuries usually occur after falls from height or as results of traffic accident 2. An increasing number of intra-articular calcaneal fractures are being reported in children 3. The management of displaced intra-articular calcaneal fractures (DIACFs) in children remains challenging and controversial 4,5. Most surgeons have favored nonoperative treatment of these fractures in the past 6,7. However, it is hard to obtain excellent anatomical reduction, especially for displaced calcaneal fractures 5. Moreover, other problems also occur, such as loss of calcaneal height, increased calcaneal width and uneven articular surface, all leading to calcaneal malunion and traumatic arthritis 5. An increasing number of studies have shown a trend toward better functional outcomes in the operatively managed groups than those treated nonoperatively 8,9.

also occur, such as loss of calcaneal height, increased calcaneal width and uneven articular surface, all leading to calcaneal malunion and traumatic arthritis 5. An increasing number of studies have shown a trend toward better functional outcomes in the operatively managed groups than those treated nonoperatively 8,9. Currently, open reduction and internal fixation through the lateral L-shape extensile incision has been considered as the gold standard surgical therapy for calcaneal fractures 10. This approach provides a large view to expose the fractures, allowing accurate reduction of the deformed posterior facet and convenient placement of the plate to achieve a stable fixation 11. However, the high incidence (~30%) of complications associated with this approach, including wound dehiscence and deep infection, remains a non-negligible problem 12. In addition, the conventional plate is difficult to insert, and wide exposure should be avoided for children with small and skeletally immature calcaneus. In an attempt to lower the complication rate, various minimally invasive techniques have been introduced recently 13,14. As one of the most widely applied minimally invasive techniques, the sinus tarsi approach has the advantage of direct visualization of the posterior articular facet and fewer wound-related complications 15.

ttempt to lower the complication rate, various minimally invasive techniques have been introduced recently 13,14. As one of the most widely applied minimally invasive techniques, the sinus tarsi approach has the advantage of direct visualization of the posterior articular facet and fewer wound-related complications 15. To further shorten the operative time and decrease the wound complication rate, we introduced, since 2010, percutaneous Kirschner wires (K-wires) fixation and allogeneic bone grafting following open reduction through the sinus tarsi approach. A preliminary clinical study carried out by our team found that this technique reduced subtalar joint stiffness and a reduced wound complication rate, while improving the patients’ satisfaction. In this study, we retrospectively analysis the data and report the functional and radiological outcomes of surgical treatment of intra-articular calcaneal fractures in children. Patients and methods Our Institutional Review Board approved the protocol for this project. We retrospectively reviewed the medical records and radiographs of 30 children who underwent open reduction for displaced, intra-articular calcaneal fractures at our institution between January 2010 and January 2015. The patients reviewed were younger than 15 years at the time of injury. We excluded cases of avulsion of the apophysis and fractures of the anterior process. Open calcaneal fractures and Sanders type I fractures were also excluded.

-articular calcaneal fractures at our institution between January 2010 and January 2015. The patients reviewed were younger than 15 years at the time of injury. We excluded cases of avulsion of the apophysis and fractures of the anterior process. Open calcaneal fractures and Sanders type I fractures were also excluded. Preoperative management After admission, the affected feet of the patients were fixed using plaster and elevated to relieve the swelling of the soft tissues. All patients were evaluated using preoperative calcaneal radiographs and computed tomography scans and three-dimensional reconstruction of the injured foot. All fractures were subsequently classified according to the Sanders classification 16 (Table 1). All patients underwent general anesthesia. The patient was placed in a supine position on a radiolucent operating table and a tourniquet was applied. Hip and knee joint were flexed. Table 1 Patients’ demographics Surgical technique We used a straight incision directly over the sinus tarsi which started ~1 cm below the lateral malleolus and continued toward the calcaneocuboid joint for ~5 cm. Dissection was carried carefully proximally to avoid injuring the calcanofibular ligament and the sural nerve. Peroneus longus and brevis tendon were pulled over and the posterior subtalar articular surface and fracture line were exposed.

ow the lateral malleolus and continued toward the calcaneocuboid joint for ~5 cm. Dissection was carried carefully proximally to avoid injuring the calcanofibular ligament and the sural nerve. Peroneus longus and brevis tendon were pulled over and the posterior subtalar articular surface and fracture line were exposed. Intra-articular hematoma and debris were eliminated through the incision. After moderate exposure of the lateral wall of the calcaneus, a periosteum elevator was inserted into the fractures and placed under the posterior articular surface. Calcaneal tuberosity was pushed backward and the posterior articular surface was pushed upward to correct the length and height of calcaneus with the help of the periosteum elevator. A 2.0- or 2.5-mm K-wire was inserted into the calcaneus tuberosity followed by longitudinally extending it to the anterior part of the calcaneus to correct the Bohler’s angle. Bone defect region was filled with allograft bone. Once satisfactory anatomical reduction of the calcaneus was achieved, 1–3 K-wires were inserted from the tuberosity to the anterior part of the calcaneus in different directions to fix the fracture line (Fig. 1). After confirming the height and width of the calcaneus, Bohler’s angle and Gissane’s angle were reduced under C-arm fluoroscopy, rubber drains were inserted, and the incision was closed in layers followed by compression bandaging.

anterior part of the calcaneus in different directions to fix the fracture line (Fig. 1). After confirming the height and width of the calcaneus, Bohler’s angle and Gissane’s angle were reduced under C-arm fluoroscopy, rubber drains were inserted, and the incision was closed in layers followed by compression bandaging. Fig. 1 Radiographic evaluations of a 7-year-old male patient with Sanders type III calcaneal fracture. Preoperative computed tomography scans show significantly reduced Bohler’s angle and Gissane’s angle (a–c). Postoperative radiograph shows significant correction of Bohler’s angle and Gissane’s angle (d). Postoperative management and evaluation Postoperatively, the affected limbs of patients were elevated to minimize swelling and immobilized by plaster cast. Prophylactic antibiotics were given to prevent surgical site infection. Rubber drainage was extracted after 48 h, and incision was cleaned every 2 days until sutures were removed ~2 weeks after surgery. Each patient received follow-up at 1, 3, 6, and 12 months postoperatively. K-wires and cast were removed after 1 month postoperatively when the fracture was healed clinically, and full weight-bearing was not allowed until bony union was confirmed on radiographs, which was ~3 months postoperatively.

weeks after surgery. Each patient received follow-up at 1, 3, 6, and 12 months postoperatively. K-wires and cast were removed after 1 month postoperatively when the fracture was healed clinically, and full weight-bearing was not allowed until bony union was confirmed on radiographs, which was ~3 months postoperatively. Radiographs were taken postoperatively to measure the calcaneal anatomical parameters, including Bohler’s angle and Gissane’s angle. Clinical functional outcomes were graded using the American Orthopedic Foot and Ankle Society scale at the 12-month follow-up 17. The postoperative wound-related complications were also recorded. Statistical analysis All data are expressed as mean±SD and analyzed by SPSS statistical package program (version 19.0; SPSS Inc., Chicago, Illinois, USA). The preoperative and postoperative calcaneal anatomical parameters were compared by paired t-test, and χ2-test was performed to compare the functional outcomes between the different Sanders classification. Fisher’s exact test was used instead when the expected frequency was less than 5. The level of significance was set at P less than 0.05.

ostoperative calcaneal anatomical parameters were compared by paired t-test, and χ2-test was performed to compare the functional outcomes between the different Sanders classification. Fisher’s exact test was used instead when the expected frequency was less than 5. The level of significance was set at P less than 0.05. Results Of the 30 patients treated since 2010, 24, including one who had bilateral fractures, were available for interview and examination. The mean age was 9.8 years (range: 5.6–14.5 years). The fractures were presented in three girls and 21 boys. The mechanisms of injury included falls from over 2 m in 17 patients, falls from below 2 m in three, and severe mangling injuries from car accidents in four. According to the Sanders classification, the fracture patterns of these 25 intra-articular injuries could be further characterized as six Sanders type II, 14 Sanders type III, and five Sanders type IV. All fractures were closed. Average postoperative clinical and radiological follow-up was 19 months (range: 12.0–30.0 months) (Table 1). The average time from initial injury to operation was 3 days (range: 2.0–5.0 days) in this study. The mean hospital stay was 10 days (range: 8.0–12.0 days), and the mean operative time was 75 min (range: 65–90 min). No deep infection and wound necrosis occurred. One patient had superficial infection around K-wires, which was resolved by dressings and oral antibiotics administered for 3 days.

.0 days) in this study. The mean hospital stay was 10 days (range: 8.0–12.0 days), and the mean operative time was 75 min (range: 65–90 min). No deep infection and wound necrosis occurred. One patient had superficial infection around K-wires, which was resolved by dressings and oral antibiotics administered for 3 days. Marked improvements in the Bohler’s angle and Gissane’s angle were noticed in all patients (Table 2). The mean preoperative and postoperative Bohler’s angles were 17.1°±10.7° and 35.9°±6.7° (P<0.05) in Sanders type II fractures, 14.4°±11.5° and 34.7°±8.5° (P<0.05) in type III, and 9.3°±9.7° and 35.1°±4.9° (P<0.05) in type IV, respectively. The mean preoperative and postoperative Gissane’s angles were 102.6°±11.5° and 125.7°±7.8° (P<0.05) in Sanders type II fractures, 101.7°±9.1° and 117.5°± 10.8° (P<0.05) in type III, and 104.7°±5.1° and 122.8°±9.1° (P<0.05) in type IV, respectively. No case of loss of reduction was noticed at the follow-up (Fig. 2). Table 2 Radiographic results for the three groups Fig. 2 Radiographic evaluations of a 13-year-old female patient with Sanders type III calcaneal fracture. Preoperative computed tomography scans show significantly reduced Bohler’s angle and Gissane’s angle (a, b). Postoperative radiographs show anatomical reduction of the subtalar articular surface and significant correction of Bohler’s angle and Gissane’s angle (c, d). Three-month follow-up radiographs show no obvious decrease in Bohler’s angle and Gissane’s angle (e, f).

antly reduced Bohler’s angle and Gissane’s angle (a, b). Postoperative radiographs show anatomical reduction of the subtalar articular surface and significant correction of Bohler’s angle and Gissane’s angle (c, d). Three-month follow-up radiographs show no obvious decrease in Bohler’s angle and Gissane’s angle (e, f). All patients were free of pain and had a normal range of motion in the subtalar joint (Fig. 3). Gait was not impaired. No secondary arthrosis has been observed and no patient required orthopedic shoes. According to the American Orthopedic Foot and Ankle Society scoring system, the mean scores of type II fractures were 92.7± 2.1, type III 90.2±1.8, and type IV 89.7±2.7 at the latest follow-ups. We did not find statistically significant differences among these three groups (P>0.05) (Table 3). Fig. 3 Radiographic and clinical evaluations of the 13-year-old female patient with Sanders type III calcaneal fracture in Fig. 2. The 12-month follow-up photographs show no obvious decrease in Bohler’s angle and Gissane’s angle and normal range of motion in the subtalar joint. Table 3 The outcome according to American Orthopedic Foot and Ankle Society scale for the three groups

Fig. 3 Radiographic and clinical evaluations of the 13-year-old female patient with Sanders type III calcaneal fracture in Fig. 2. The 12-month follow-up photographs show no obvious decrease in Bohler’s angle and Gissane’s angle and normal range of motion in the subtalar joint. Table 3 The outcome according to American Orthopedic Foot and Ankle Society scale for the three groups Discussion Fractures of the calcaneus make up ~2% of all fractures and are the commonest fracture of the tarsal bones 1. The optimal management of DIACFs remains controversial 18. Many surgeons have favored nonoperative treatment of these fractures; however, Thermann et al. 19 reported arthritic changes in the subtalar joint in 65.2% of type II fractures and in 81.7% of type III and type IV fractures, and radiological follow-up evaluation showed arthritic changes in the subtalar joint in 44% of cases. Certain randomized controlled trials supported surgical treatment can better reconstruct the anatomy of the calcaneus, lower the subtalar fusion rate, and offer protection against early subtalar arthrodesis in DIACFs 9.

nd radiological follow-up evaluation showed arthritic changes in the subtalar joint in 44% of cases. Certain randomized controlled trials supported surgical treatment can better reconstruct the anatomy of the calcaneus, lower the subtalar fusion rate, and offer protection against early subtalar arthrodesis in DIACFs 9. Schneidmueller et al. 20 aimed to restore articular surface through surgery to prevent early arthritis. They reported surgical treatment is recommended when the disruption of the three subtalar facets is more than 2 mm. Depending on the outcome score used, an overall good to excellent result was reached in 90% of cases, without wound complication. Pickle et al. 21 reported a group of seven pediatric intra-articular calcaneal fractures surgically treated with internal fixation. In our study, all patients were free of pain and had a normal range of motion of the subtalar joint. In agreement with these authors, reconstruction of the joint alignment and satisfactory functional outcomes were achieved by surgery for DIACFs in our study.

alcaneal fractures surgically treated with internal fixation. In our study, all patients were free of pain and had a normal range of motion of the subtalar joint. In agreement with these authors, reconstruction of the joint alignment and satisfactory functional outcomes were achieved by surgery for DIACFs in our study. Closed reduction and percutaneous fixation have been reported to significantly reduce the incidence of wound healing complications 22. However, this treatment seems to be more suitable for cases with moderately displaced fracture or noncomminuted calcaneus fracture because of the limited fixation strength. We found the articular surface of the immature calcaneum is difficult to restore by closed reduction because of their high resilience and low rigidity. Surgical treatment with open reduction and internal fixation is increasingly recommended in the literature 11. Alternative surgical approaches to the calcaneus include the lateral L-shape, Palmer approach, and Kocher approach 23,24. The lateral L-shape extensile incision provides excellent exposure of the fractures and allows accurate reduction of the deformed posterior facet and convenient placement of the plate to achieve a stable fixation. However, open reduction and internal fixation through this approach comes with various complications, among which, the wound complications, including edge necrosis, dehiscence, hematoma, and/or deep infection, remain a major concern 23.

posterior facet and convenient placement of the plate to achieve a stable fixation. However, open reduction and internal fixation through this approach comes with various complications, among which, the wound complications, including edge necrosis, dehiscence, hematoma, and/or deep infection, remain a major concern 23. Many minimally invasive techniques were developed to overcome this disadvantage 8,14. Xia et al. 14 reported shorter surgical times and lower wound complications using the minimally invasive sinus tarsi approach compared with an extended L-shaped lateral approach. They reported the overall rate of wound complications was 6% in the patients treated with sinus tarsi approach. In this study, an incision (4–5 cm) was made from the level of the tip of the lateral malleolus and continued toward the calcaneocuboid joint. K-wires were used to fix the fractures and the incision was only used to assist the repositioning and filling of the bone graft. There is no need for extensive periosteal dissection to provide the internal fixation space, which is necessary in operation for adults. We need to reduce the length of the incision and the dissection of the periosteum during the operation. This type of small incision greatly decreases the effect on local skin blood supply. Only one patients in our study had superficial infection around K-wires, which was resolved by dressings and oral antibiotics administered for three days. Compared with conventional treatment, sinus tarsi approach substantially decreased the risk of wound complications. In addition, K-wires can be inserted percutaneously, thus eliminating the need for the creation of a complete subcutaneous tunnel, which can avoid injuries to the peroneal tendons and the sural nerve.

ys. Compared with conventional treatment, sinus tarsi approach substantially decreased the risk of wound complications. In addition, K-wires can be inserted percutaneously, thus eliminating the need for the creation of a complete subcutaneous tunnel, which can avoid injuries to the peroneal tendons and the sural nerve. Stulik et al. 25 reported 13 (4.5%) cases of loss of reduction in patients treated with closed reduction and percutaneous fixation using K-wires. In our study, allogeneic bone was used to fill the calcaneal defect followed by open reduction. Although bone grafts do add mechanical support, and may stimulate quicker fracture healing, the calcaneus heals rapidly with reconstitution of the cancellous bone within 3 months after surgery without loss of reduction.

res. In our study, allogeneic bone was used to fill the calcaneal defect followed by open reduction. Although bone grafts do add mechanical support, and may stimulate quicker fracture healing, the calcaneus heals rapidly with reconstitution of the cancellous bone within 3 months after surgery without loss of reduction. Conclusion Good treatment results were obtained in all of the fractured calcaneus in this study. Our excellent results may also be attributable to the sinus tarsi approach, percutaneous K-wires fixation, and allogeneic bone graft. There are several advantages such as (i) the articular surface was well exposed, and the restoration can be performed under direct vision; (ii) sinus tarsi approach facilitates bone graft, which can increase bone strength and reduce the risk of collapse; and (iii) it is easy to insert and remove the K-wires. Moreover, there is no need for extensive periosteal dissection to provide the internal fixation space. No additional operation is required to remove the K-wires. However, there are still some limitations in this technique. This minimally invasive sinus tarsi approach with K-wires fixation technique is still difficult for exposure and restoration of fractured fragments of the medial wall of the calcaneus. The long-term complications of calcaneal fractures in children include osteoarthritis and loss of calcaneal height in the process of development. Further investigation with large sample size and longer follow-up time is still needed to obtain a more precise clinical efficacy.

gments of the medial wall of the calcaneus. The long-term complications of calcaneal fractures in children include osteoarthritis and loss of calcaneal height in the process of development. Further investigation with large sample size and longer follow-up time is still needed to obtain a more precise clinical efficacy. Acknowledgements All authors have participated sufficiently in this work concerning conception and design of this study, drafting the article, critical revision for important intellectual content, and final approval. The authors are grateful to Tao Tong from the Rockefeller University for the helpful advice about expression. This work was supported by Nature Science Foundation from the Science and Technology Department of Hubei Province (grant number: 2017CFB385) and Clinical Medical Research Project from the Health and Family Planning Commission of Wuhan Municipality (grant number: WX17Q20). Conflicts of interest There are no conflicts of interest.

Introduction A brain lesion during gestation or in the immature brain may result in cerebral palsy (CP), which is characterized by motor impairments but also includes many other symptoms, depending on the timing, location, and extent of the brain damage 1,2. Asymmetry of the injury typically results in unilateral CP, with one arm and leg more affected than the other. An acquired brain injury (ABI), occurring after 5 years of age, may also result in spastic hemiplegia 3. In both groups, contractures are common, with progressive stiffening and shortening of the musculotendinous unit 4. In a population based study on 771 children with CP, about one third developed contractures which became significant at 4 years of age regarding wrist extension with straight fingers, and at 7 years of age regarding supination and elbow extension 5. There is one study on the effect of lower limb multilevel surgery with a comparison with a control group, albeit treated with botulinum toxin and physiotherapy. It was found that the operated group showed improved gait and range of motion, whereas the unoperated group showed deterioration 6. In the upper limb, the muscles most prone to contracture formation are the biceps brachii, the pronator teres (PT), the flexor carpi ulnaris (FCU), the flexor carpi radialis, superficial and deep finger flexors, and the adductor pollicis (AddP) muscles. Cocontraction and poor selective muscle control also contribute toward the typical position with the elbow and wrist flexed and the forearm pronated 7. When the child runs or is excited, the flexed position of the arm is aggravated further, which impairs the function and appearance of the arm even more. Impairment of one hand affects the activity in daily life as we are typically doing most things with two hands. Persons with CP or ABI will naturally use the better hand the most, but the paretic hand can be very useful as a helper hand. The usefulness of the paretic hand can be measured with the Assisting Hand Assessment (AHA), and the score at 18 months can predict the score at 12 years of age 8.

ng most things with two hands. Persons with CP or ABI will naturally use the better hand the most, but the paretic hand can be very useful as a helper hand. The usefulness of the paretic hand can be measured with the Assisting Hand Assessment (AHA), and the score at 18 months can predict the score at 12 years of age 8. Progressive flexion of the wrist and elbow, adduction of the thumb, and a fixed pronation of the forearm will make it more difficult to approach, hold, and manipulate objects, and body appearance is also affected. This can be treated surgically by individually tailored tendon lengthenings and tendon transfers. Even though tendon transfer for the spastic hand was first described more than half a century ago 9, the evidence for benefits of upper limb surgery in a long-term perspective is low and there are few available studies 10. Short-term follow-up of surgery has shown a functional result 11,12 and the effect has even been observed in bimanual activities, measured with the AHA 13. The aim in this study was to investigate whether short-term improvements in active range of motion and bimanual hand function (AHA) are maintained at a 9-year follow-up by investigating the same participants as in a previous study 13. In that study, we found that the operated upper limb was more useful in bimanual activities and that the range of motion was improved. We also wanted to know how the appearance of the hand was appreciated by the child.

re maintained at a 9-year follow-up by investigating the same participants as in a previous study 13. In that study, we found that the operated upper limb was more useful in bimanual activities and that the range of motion was improved. We also wanted to know how the appearance of the hand was appreciated by the child. Patients and methods This is a case series, performed at the Department of Pediatric Orthopaedic Surgery, Karolinska University Hospital. In a previous study, we examined 18 children with CP consecutively planned for hand surgery, and then performed a follow-up at 7 months 13. For 15 of the 18 children in our previous study, at least 7.5 years had passed after the first surgery, and they were contacted for a follow-up assessment. Two of the remaining 15 patients declined to participate in the follow-up. We included two more patients, no. 7 and 15, with 10 and 8.2 years of follow-up, resulting in a total of up to 15 patients. Twelve of the patients had unilateral CP, and three ABI; one had had a traumatic brain injury at age 10, one was operated for a brain tumor at age 9, and one had had a stroke at 9 years. Details on age at first surgery (11.3±3.5) and age at long-term follow-up (20.5±3.5 years) are presented in Table 1. Long-term follow-up was performed at an average of 9 years (106 months, range: 90–126 months) after surgery of the upper extremity. Table 1 Patients with 9.3 years of follow-up of hand surgery in hemiplegia

For 15 of the 18 children in our previous study, at least 7.5 years had passed after the first surgery, and they were contacted for a follow-up assessment. Two of the remaining 15 patients declined to participate in the follow-up. We included two more patients, no. 7 and 15, with 10 and 8.2 years of follow-up, resulting in a total of up to 15 patients. Twelve of the patients had unilateral CP, and three ABI; one had had a traumatic brain injury at age 10, one was operated for a brain tumor at age 9, and one had had a stroke at 9 years. Details on age at first surgery (11.3±3.5) and age at long-term follow-up (20.5±3.5 years) are presented in Table 1. Long-term follow-up was performed at an average of 9 years (106 months, range: 90–126 months) after surgery of the upper extremity. Table 1 Patients with 9.3 years of follow-up of hand surgery in hemiplegia Surgical procedures and number of surgeries performed are listed for each patient in Table 2. Each child was assessed before surgery and combinations of surgical techniques involving the elbow, wrist, and thumb were tailored individually with the main aim of improving reach, grip, release, and object manipulation. Table 2 Surgical aims, number of procedures/goal

Surgical procedures and number of surgeries performed are listed for each patient in Table 2. Each child was assessed before surgery and combinations of surgical techniques involving the elbow, wrist, and thumb were tailored individually with the main aim of improving reach, grip, release, and object manipulation. Table 2 Surgical aims, number of procedures/goal After surgery, all patients wore a long-arm cast for 6 weeks, maintaining the arm in an overcorrected position. Following cast removal, intense training (2 months, 1–3 times/week, 1–2 h/session) was performed under the supervision of an occupational therapist (OT). Less frequent training (1 time/week, 1–2 h/session) was continued up until 6 months after surgery. During the initial 6 months, the children were also encouraged to use the hand in their daily activities through an individually designed home-based training program. No supervised training intervention was maintained thereafter. According to routine, a post-OP AHA and clinical examination was performed about 6 months after surgery. All clinical assessments were performed by the same two OTs (C.L.E. and A.L.). The AHA films were scored by an OT not involved in the treatment or follow-up (B.M.Z.).

After surgery, all patients wore a long-arm cast for 6 weeks, maintaining the arm in an overcorrected position. Following cast removal, intense training (2 months, 1–3 times/week, 1–2 h/session) was performed under the supervision of an occupational therapist (OT). Less frequent training (1 time/week, 1–2 h/session) was continued up until 6 months after surgery. During the initial 6 months, the children were also encouraged to use the hand in their daily activities through an individually designed home-based training program. No supervised training intervention was maintained thereafter. According to routine, a post-OP AHA and clinical examination was performed about 6 months after surgery. All clinical assessments were performed by the same two OTs (C.L.E. and A.L.). The AHA films were scored by an OT not involved in the treatment or follow-up (B.M.Z.). The AHA measures and describes how efficiently people with unilateral hand-dysfunction spontaneously use their affected hand in bimanual activities. It is a criterion-referenced test proved to be valid, reliable, and responsive to change for children 18 months to 18 years of age 14. The assessment is a standardized videotaped 10–15 min long semi-structured play session. Specifically chosen toys requiring the use of both hands are used. For adolescents, a semi-structured activity is used, including opening and wrapping a present and writing and placing a note in an envelope. It includes 22 items, describing different actions of the assisting hand, each scored on a four-point rating scale. The raw score range from 22 to 88 (low to high ability). The raw sum score is converted into the logit-based 0–100 AHA-unit scale on the basis of Rasch measurement analysis. The smallest detectable difference for the AHA has been calculated to five AHA-units; this means that on an individual level, a change of 5 U or more can be considered a true measurement change, also clinically and functionally detectible 15. We based the scoring on the AHA 4.4 version 14,16.

asch measurement analysis. The smallest detectable difference for the AHA has been calculated to five AHA-units; this means that on an individual level, a change of 5 U or more can be considered a true measurement change, also clinically and functionally detectible 15. We based the scoring on the AHA 4.4 version 14,16. Active range of motion Active range of motion of the elbow, forearm, and wrist was measured using a goniometer. On the basis of the principle that a neutral position equals 0°, measurements were taken from pronation to supination of the forearm and from flexion to extension of the wrist and elbow. For the elbow, limitation of extension to 0° is stated as an extension deficit (American Academy of Orthopaedic Surgeons, 1965). Zancolli classification Simultaneous voluntary wrist and finger extension was classified into four groups according to Zancolli’s classification 17. In type 1, the fingers can extend while the wrist is in less than 20° flexion; in type 2A, the fingers can extend with the wrist in more than 20° of flexion and when the fingers are flexed, the wrist can extend. In type 2B, the fingers can extend when the wrist is flexed more than 20°, but the wrist cannot extend even with the fingers flexed. In type 3, neither the fingers nor the wrist can extend. Klingels et al. 18 have shown the classification to have excellent reliability.

when the fingers are flexed, the wrist can extend. In type 2B, the fingers can extend when the wrist is flexed more than 20°, but the wrist cannot extend even with the fingers flexed. In type 3, neither the fingers nor the wrist can extend. Klingels et al. 18 have shown the classification to have excellent reliability. House classification of thumb position The position of the thumb during active grasp was classified into four groups described by House 11. The evaluation was performed by observing when the patients opened their hand to actively grasp objects. House I=Adduction in the first carpometacarpal joint (CMC I) because of increased tone and/or contracture of AddP, House II=adduction in CMC I and flexion in the first metacarpal-phalangeal joint (MCP I) because of increased tone and/or contracture in AddP and flexor pollicis brevis muscles, House III=adduction in CMC I and hyperextension in MCP I and/or first interphalangeal joint, and House IV=adduction of the CMC I and flexion of MCP I and interphalangeal joint 19. House Classification of Upper Extremity Functional Use was developed to classify the use of the hemiplegic hand 19. It is a nine-graded scale, with categories ranging from no use at all (0), to spontaneous use (8). The evaluation was determined by observing how active the affected hand was in selected activities.

House classification of thumb position The position of the thumb during active grasp was classified into four groups described by House 11. The evaluation was performed by observing when the patients opened their hand to actively grasp objects. House I=Adduction in the first carpometacarpal joint (CMC I) because of increased tone and/or contracture of AddP, House II=adduction in CMC I and flexion in the first metacarpal-phalangeal joint (MCP I) because of increased tone and/or contracture in AddP and flexor pollicis brevis muscles, House III=adduction in CMC I and hyperextension in MCP I and/or first interphalangeal joint, and House IV=adduction of the CMC I and flexion of MCP I and interphalangeal joint 19. House Classification of Upper Extremity Functional Use was developed to classify the use of the hemiplegic hand 19. It is a nine-graded scale, with categories ranging from no use at all (0), to spontaneous use (8). The evaluation was determined by observing how active the affected hand was in selected activities. Appreciation of appearance of upper extremity Patients and, in the cases where the patients could not answer, their parents were contacted by telephone. The interview included four questions, presented to each patient in the same order. The patients were asked to rate their satisfaction with the appearance of their hand/arm from a scale of 1 to 10. The questions were as follows: (a) How pleased are you with the appearance of your hand and arm when you walk? (b) Can you recall how pleased you were before the surgical treatment? (c) How pleased are you with the appearance of your hand and arm when you sit down? (d) Can you recall how pleased you were before the surgical treatment? The telephone calls were made by an MD who had not taken part in the care of the patient at any stage (F.v.W.).

l how pleased you were before the surgical treatment? (c) How pleased are you with the appearance of your hand and arm when you sit down? (d) Can you recall how pleased you were before the surgical treatment? The telephone calls were made by an MD who had not taken part in the care of the patient at any stage (F.v.W.). Statistical analysis Nonparametric methods were used to define the outcome of the surgery. Results were compared using Friedman’s test for repeated measures, with Dunn’s post-hoc test for multiple comparisons between groups. The short-term AHA was missing for patient no. 7 and 15. As their AHA at other time-points during their childhood and adolescence were very stable, we estimated the short-term AHA as the mean between pre-OP and long-term follow-up. GraphPad Prism 7 was used for calculations (GraphPad Software, La Jolla, California, USA). For all statistical tests, a probability of P value of less than 0.05 was adopted as the criteria for determining significant differences. Values in the text are expressed as the median, the minimum, and the maximum, apart from age, which is expressed as mean. Median and interquartile range were used for all figures. No statistics were performed on the categorical data, Zancolli classification, House classification for hand function, House thumb classification, or how the child appreciated the appearance of the affected hand.

imum, apart from age, which is expressed as mean. Median and interquartile range were used for all figures. No statistics were performed on the categorical data, Zancolli classification, House classification for hand function, House thumb classification, or how the child appreciated the appearance of the affected hand. Results There was a wide range of hand function in the children included, evidenced by a large span in the AHA units at pre-OP, ranging from 7 to 64, with a median of 50 U. At the post-OP follow-up at about 7 months, AHA had improved to 52 U, range 24–67 (main effect of time P=0.0243), but declined to 49 U, range 12–67 at the 9-year follow-up (Fig. 1a). However, different items of the AHA did not change uniformly (Fig. 1b). The AHA test items ‘calibrates’, ‘stabilizes by grip’, ‘flow in performance’, ‘orients objects’, and ‘approaches objects’ were still improved compared with before surgery, whereas ‘manipulates’, ‘reaches’, ‘coordinates’, ‘initiates use,’ and ‘changes strategy’ had become worse. Fig. 1 (a) Assisting Hand Assessment (AHA) score (logits converted into a 0–100 scale) pre, post (6 months), and follow-up (9 years). *Significantly different pre to post, P<0.05. Data are presented as median and interquartile range. (b) Changes in raw scores between the assessments before surgery and at the 9-year follow-up, distributed by the separate Assisting Hand Assessment test items with mean values for the group. The items are ordered and numbered in accordance with the hierarchy of difficulty established by the Rasch analysis with the easiest item at the bottom.

es between the assessments before surgery and at the 9-year follow-up, distributed by the separate Assisting Hand Assessment test items with mean values for the group. The items are ordered and numbered in accordance with the hierarchy of difficulty established by the Rasch analysis with the easiest item at the bottom. Active elbow extension deficit was present in 12 out of 15 children before surgery, median 18°, range 0°–35°(Fig. 2a, left). Biceps tendon lengthening was performed in nine of the children and resulted in a significant improvement of the active extension of the elbow at 7 months after OP (Fig. 2a, middle), which was largely maintained up until 9 years of follow-up (pre-OP extension deficit 30° (range: 5°–35°), 7-month post 15° (range: 0°–25°), and 9-year follow-up 15° (range: 0°–30°), pre-OP to 7-month post, and post–to 9-year follow-up P=0.0267. However, in the children who did not undergo surgery of the elbow flexors, there was a significant reduction of the active elbow extension at the 9-year long-term follow-up (pre-OP extension deficit 5° (range: 0°–25°), post-OP 7 months 0° (range: 0°–25°) and follow-up 9 years 16.5° (range: 0°–30°), and 7-month-post-OP to 9-year-follow-up P=0.0424 (Fig. 2a, right). As shown in Table 2, all children underwent surgical procedures aiming to increase active supination. Active supination increased significantly following surgery, pre-OP to 7 months post-OP P=0.0140, and seemed to remain improved at the 9-year follow-up, but was not statistically significant [pre 0° (range: −80° to 90°), post-OP 60° (range: −50° to 90°), and 9-year follow-up 55° (range: −90° to 90°)] (Fig. 2b). All children, except one, were operated on with the aim of improving wrist extension using a variety of surgical techniques (Table 2). Surgery improved active wrist extension more than 50° when evaluated 7 months post-OP (P=0.0185) and was still significantly improved at the 9-year follow-up (P=0.0078) compared with pre-OP [pre −10° (range: −80° to 60°), post 50° (range: −10° to 65°), and follow-up 55° (range: −70° to 70°)] (Table 3).

iques (Table 2). Surgery improved active wrist extension more than 50° when evaluated 7 months post-OP (P=0.0185) and was still significantly improved at the 9-year follow-up (P=0.0078) compared with pre-OP [pre −10° (range: −80° to 60°), post 50° (range: −10° to 65°), and follow-up 55° (range: −70° to 70°)] (Table 3). Fig. 2 (a) Extension defect (deg.) of the elbow joint in left: all included patients (n=15) middle: all patients undergoing biceps lengthening (n=9) right: all nonsurgery cases (n=6). (b) Active supination (deg.) of the wrist in all included patients (n=15). (c) Active extension (deg.) of the wrist in all included patients (n=15). *,**Denotes significantly different from the time point indicated by the square bracket, P<0.05, P<0.01 respectively. For all graphs, data are presented as median and interquartile range. Table 3 Active range of motion Results of the Zancolli’s classification of active wrist and finger extension are shown in Table 4. Compared with pre-OP, at the 9-year follow-up, for four it was more difficult to extend the wrist and the fingers, for another four it was easier to extend wrist and fingers, while there was no change in seven patients. Patient 11 was operated on with a proximal row carpectomy and wrist fusion 4 years after the first surgery. He, for obvious reasons, could not extend the wrist dorsally, but with the wrist fixed at 0°, he could extend straight fingers fully, and was therefore classified as Zancolli 1. Table 4 House functional class, House classification of thumb and Zancolli classification

Results of the Zancolli’s classification of active wrist and finger extension are shown in Table 4. Compared with pre-OP, at the 9-year follow-up, for four it was more difficult to extend the wrist and the fingers, for another four it was easier to extend wrist and fingers, while there was no change in seven patients. Patient 11 was operated on with a proximal row carpectomy and wrist fusion 4 years after the first surgery. He, for obvious reasons, could not extend the wrist dorsally, but with the wrist fixed at 0°, he could extend straight fingers fully, and was therefore classified as Zancolli 1. Table 4 House functional class, House classification of thumb and Zancolli classification The operated hand was also classified according to the House Classification of Upper Extremity Functional Use (Table 4). Compared with pre-OP, at the 9-year follow-up, five patients still had a passive grip (House 1–3), seven still had an active grip (House 4–6), and one patient still used the hand spontaneously. In one patient, passive function of the hand had become active (House 3 to >5), and in one patient, from active to spontaneous use (House 5 to >7).

at the 9-year follow-up, five patients still had a passive grip (House 1–3), seven still had an active grip (House 4–6), and one patient still used the hand spontaneously. In one patient, passive function of the hand had become active (House 3 to >5), and in one patient, from active to spontaneous use (House 5 to >7). The patients who were treated for thumb in palm by different tendon transfers and releases are reported in Table 2. On comparing the House classification of thumb position pre-OP with the 9-year follow-up, six had a more functional thumb position, three patients had a more dysfunctional position, and there was no change in 3 (three patients had missing values at pre-OP (Table 4). Some of the patients had required more than one operation for the thumb to have an acceptable position at the initiation of a grip. Twelve of the 15 patients took part in the telephone interview. Ten reported that they were now more satisfied, at the 9-year follow-up, with the appearance of the hand and arm when walking (Fig. 3a). Similarly, 9/12 patients were more satisfied with the appearance of the hand and arm when sitting down now compared with before the surgery (Fig. 3b). Fig. 3 (a) Patient-rated appreciation (0–10 scale) of the upper extremity before surgery and at the 9-year follow-up during walking. (b) Patient-rated appreciation (0–10 scale) of the upper extremity before surgery and at the 9-year follow-up while sitting down. n=12 for both graphs.

The patients who were treated for thumb in palm by different tendon transfers and releases are reported in Table 2. On comparing the House classification of thumb position pre-OP with the 9-year follow-up, six had a more functional thumb position, three patients had a more dysfunctional position, and there was no change in 3 (three patients had missing values at pre-OP (Table 4). Some of the patients had required more than one operation for the thumb to have an acceptable position at the initiation of a grip. Twelve of the 15 patients took part in the telephone interview. Ten reported that they were now more satisfied, at the 9-year follow-up, with the appearance of the hand and arm when walking (Fig. 3a). Similarly, 9/12 patients were more satisfied with the appearance of the hand and arm when sitting down now compared with before the surgery (Fig. 3b). Fig. 3 (a) Patient-rated appreciation (0–10 scale) of the upper extremity before surgery and at the 9-year follow-up during walking. (b) Patient-rated appreciation (0–10 scale) of the upper extremity before surgery and at the 9-year follow-up while sitting down. n=12 for both graphs. Discussion This long-term follow-up of hand surgery in children with spastic hemiplegia shows that a progressive malpositioning of the hand can be halted with tendon lengthenings and transfers. The short-term advances in bimanual function return to pre-OP status after some years of self-selected use of the hand.

Fig. 3 (a) Patient-rated appreciation (0–10 scale) of the upper extremity before surgery and at the 9-year follow-up during walking. (b) Patient-rated appreciation (0–10 scale) of the upper extremity before surgery and at the 9-year follow-up while sitting down. n=12 for both graphs. Discussion This long-term follow-up of hand surgery in children with spastic hemiplegia shows that a progressive malpositioning of the hand can be halted with tendon lengthenings and transfers. The short-term advances in bimanual function return to pre-OP status after some years of self-selected use of the hand. The usefulness of the assisting hand, tested with the AHA, was first improved at 7 months, and then returned to the pre-OP level at 9 years, similar to both CP and ABI. In our previously published study with short-term follow-up 6 months after hand surgery, we found that the AHA test items ‘grasps’, ‘holds’, ‘stabilizes by grip’, ‘readjusts grip’, ‘calibrates’, ‘manipulates,’ and ‘flow in performance’ improved the most, probably because of diminished wrist flexion and pronation, which is a desired result of, for example, a Green transfer (FCU to ECRB or EDC) (Fig. 1a) 13. A previous study has shown a correlation between AHA scores and active supination and dorsal extension of the wrist 20. As FCU to ECRB (green) transfer and a PT release or rerouting improves dorsal extension of the wrist and supination, there seems to be a potential for improvement of the AHA post-OP. However, other studies have not been able to show an effect of hand surgery on AHA at follow-up. A multicenter study comparing (a) FCU with ECRB tendon transfer+PT release+release of AddP+rerouting of EPL, (b) standardized botulinum toxin treatment of FCU, PT, AddP, and (c) standard therapy regimen found no significant change in AHA in any group after 12 months 21. In our study, the procedures were not standardized, but individually tailored for each child after analyzing hand and arm activity, and after assessing true contractures during the surgery. In this way, the surgery could be optimized for each individual.

und no significant change in AHA in any group after 12 months 21. In our study, the procedures were not standardized, but individually tailored for each child after analyzing hand and arm activity, and after assessing true contractures during the surgery. In this way, the surgery could be optimized for each individual. Another contributing factor for our short-term AHA improvement could be that during the first 6 months after the surgery, training of the operated assisting hand was prioritized. In our regimen, the training is home based, with visits to the OT once or twice a week after the casting period, which sums up to about 20 sessions. At home and in school, the children do their normal activities, but with a special focus on the operated hand. The aim is that training of everyday tasks is repeated enough so that functional goals can withstand distractions, which probably has an impact on the short-term AHA improvement.

sums up to about 20 sessions. At home and in school, the children do their normal activities, but with a special focus on the operated hand. The aim is that training of everyday tasks is repeated enough so that functional goals can withstand distractions, which probably has an impact on the short-term AHA improvement. At our 9-year follow-up after the first surgery, the mean total AHA score had returned to the pre-OP status, proposing that in the long run, the amount of daily automatic use of the most affected hand is more important for bimanual function than is the muscle balance around the wrist. When the study started, the patients were children or adolescents who trained at the initiative of their parents or teachers, but at the long-term follow-up they had become adults, and could decide themselves whether they wanted to train their affected hand or not. Most had little motivational drive to train their hand and arm specifically. Automation of trained tasks declined, increasing the risk for dual-task interference 22–25. The AHA test items that had deteriorated the most at long term were ‘manipulates’, ‘coordinates’, ‘initiates use’, and ‘changes strategy’, which are items that reflect inherent motor control and attentional demands the most. The item ‘reaches’ had also deteriorated, either because the arm is longer when children had become adults and they therefore do not have to reach as much, or that the contracture formation of the elbow flexors had progressed. However, the items ‘stabilizes by grip’, ‘calibrates’, and ‘flow in performance’ were still improved compared with pre-OP, implying that the improved position of the wrist makes gripping easier (Fig. 1b). However, as yet, there is no study on the natural course of the AHA into adulthood in patients with CP or ABI, only till 12 years of age, before many severe contractures have developed 8. Progressive contractures of the wrist, thumb, and pronators will put the hand in a dysfunctional position, which may affect the AHA negatively 20. A small decline in the AHA in absolute numbers 9 years after surgery may thus then instead be better than an assumed natural course of deterioration of the AHA score.

ped 8. Progressive contractures of the wrist, thumb, and pronators will put the hand in a dysfunctional position, which may affect the AHA negatively 20. A small decline in the AHA in absolute numbers 9 years after surgery may thus then instead be better than an assumed natural course of deterioration of the AHA score. We know that the contracture formation starts early in the lower limb, long before evident contractures can be seen, and also in muscles with a minimal increase in tone 4,26. A similar phenomenon occurs in the upper extremity, with progressive adduction of the shoulder, flexion of the elbow, and flexion and pronation of the wrist. Wrist flexors have been shown to be stiffer and thinner compared with typically developed patients 27, and the biceps brachii muscle has been shown to have more ECM around the fiber bundles, probably increasing stiffness, and fewer satellite cells and less ribosomal RNA as a sign of reduced growth 28.

ion of the wrist. Wrist flexors have been shown to be stiffer and thinner compared with typically developed patients 27, and the biceps brachii muscle has been shown to have more ECM around the fiber bundles, probably increasing stiffness, and fewer satellite cells and less ribosomal RNA as a sign of reduced growth 28. We used the FCU muscle, as a transfer in 13/15 patients as it takes away the flexing and ulnarly deviating force that it has on the wrist, whereas it enables both supination and wrist extension when it is transferred around the ulna to finger/wrist extensors. However, long-term follow-up shows that it is common for the transfer to deteriorate. It is a muscle prone to progressive shortening, myopathic changes, and stiffening 29–32. A transfer of the FCU therefore has the potential to become tighter over time, causing an extension contracture of the wrist, sometimes called a reversed contracture, and perhaps a static supination 27. Six of the 15 patients in this study needed repeated surgeries because of progressive contractures, and follow-up till adulthood is therefore recommended.

he potential to become tighter over time, causing an extension contracture of the wrist, sometimes called a reversed contracture, and perhaps a static supination 27. Six of the 15 patients in this study needed repeated surgeries because of progressive contractures, and follow-up till adulthood is therefore recommended. The transfer could also rupture or become too loose or weak, leading to a flexed wrist again 33. A strong side-to-side tendon transfer technique adopted from tetraplegia surgery could be a better choice 34. One example is patient no. 9, who, at the long-term follow-up, was again unable to extend the wrist, making it more difficult grip and hold and to see what is in the hand. A person with a relapse like this could benefit from a surgical repair of the transfer or taking another muscle as a donor to the ECRB, or a proximal row carpectomy and a wrist fusion would also lead to a relative lengthening of the contracted finger flexors and a stable wrist. With an effortlessly stable wrist, more focus can be placed on flexing and extending the fingers during grip and manipulation of objects.

another muscle as a donor to the ECRB, or a proximal row carpectomy and a wrist fusion would also lead to a relative lengthening of the contracted finger flexors and a stable wrist. With an effortlessly stable wrist, more focus can be placed on flexing and extending the fingers during grip and manipulation of objects. Even though the AHA had returned to the pre-OP level, many patients appreciated that the surgery had been performed as they liked the appearance of the hand. Less flexion of the arm during gait has been shown to correlate to better self-esteem and sense of coherence 35, and most patients in this study appreciated the look of the arm when walking. General cosmesis of the hand when sitting is also a concern 36 and has been shown to be improved after hand surgery in CP 37,38. While interviewing these young adults, many stated that the function of their afflicted hand was still quite poor, but the surgery made it easier to hold the cell-phone and, for example, text with better hand, which is important in communicating with others. The hand looked better while resting in the lap, and the arm did not flex as much while walking or running, making the patients feel less different.

hand was still quite poor, but the surgery made it easier to hold the cell-phone and, for example, text with better hand, which is important in communicating with others. The hand looked better while resting in the lap, and the arm did not flex as much while walking or running, making the patients feel less different. Limitations This is a clinical study of a small number of patients with diverse impairments treated with individually tailored surgical procedures. Therefore, a prospective randomization design of different tendon transfers, etc. was not possible. The treatment methods and surgical techniques have been chosen according to the literature and clinical experience. By comparing the clinical status before surgery with both 7-month short-term and 9-year long-term follow-up, we believe that some conclusions can be made. Missing data are a common problem in clinical studies. For two patients, we replaced missing 7-month short-term AHA with the mean between pre-OP and long-term post-OP AHA (+2 and −1.5 U from pre-OP AHA, respectively), which was less than the mean increase of 6 U in the whole cohort. Our estimations of missing values did thus not exaggerate our results or conclusions. However, ideally, a multicenter randomized-controlled study should be carried out in the future. Large-scale longitudinal studies of the development of contractures and of the AHA through adolescence would also help to interpret the effect of treatment.

missing values did thus not exaggerate our results or conclusions. However, ideally, a multicenter randomized-controlled study should be carried out in the future. Large-scale longitudinal studies of the development of contractures and of the AHA through adolescence would also help to interpret the effect of treatment. Conclusion Brain lesions in children may result in deficient motor control and progressive contracture formation in the arm, leading to flexed wrist, elbow and fingers, and diminished supination. In our study, tendon transfer surgery and tendon lengthenings/releases improved bimanual function in a short-term perspective, and returned to pre-OP function in the long term. If elbow flexors had been lengthened, active elbow extension was maintained long term, but if no surgery had been performed, contracture formation progressed. The hope is that in the future, muscles prone to growth disturbance and stiffening can be identified early by, for example, ultrasound or MRI. Prophylactic treatment could then be individually tailored, and if surgery is still needed, the optimal tendon transfer for that specific patient could be planned. Acknowledgements The authors thank Anna Lundberg, OT, for performing some of the follow-up measurements, and Lena Krumlinde-Sundholm for valuable advice. Conflicts of interest There are no conflicts of interest.

Introduction Supracondylar humerus fractures, the second most common type of paediatric fractures [1], according to Gartland’s criteria, are classified as nondisplaced fractures (type I), hinged fractures with the posterior cortex intact (type II) and completely displaced fractures (type III) [2]. Supracondylar humerus fractures occur most commonly during the first decade of life, but can also occur in children older than 10 years. Several studies showed higher incidence of displacement of supracondylar humerus fractures in older children and indicated that the number and configuration of the internal fixation should be modified according to age [3]. However, very few studies on the approach and internal fixation of the supracondylar humerus fractures for the older children have been carried out. We retrospectively analysed all the data of older children aged older than 10 years with Gartland type III supracondylar humerus fractures in our clinics between January 2010 and December 2015, and reported the clinical and radiographic long-term results.

humerus fractures for the older children have been carried out. We retrospectively analysed all the data of older children aged older than 10 years with Gartland type III supracondylar humerus fractures in our clinics between January 2010 and December 2015, and reported the clinical and radiographic long-term results. Nonoperative and operative methods include long-arm plaster immobilization, axial traction applied using tape or a transolecranon pin, external fixation and percutaneous pinning with or without open reduction [4–6]. Percutaneous pinning after closed reduction is the most widely advocated technique, which usually produces a satisfactory cosmetic and functional result, and has become a standard method of treatment. However, sometimes, anatomical reduction is difficult to achieve by closed reduction when the patients with Gartland type III fracture present after a delay of a few days with swollen elbows. Also, the lack of fluoroscopy and radiation protective equipment precludes successful closed management. Residual dislocation may lead to cubitus varus, especially in older children [7]. At the same time, iatrogenic ulnar nerve injury may occur during the procedure of cross pinning [8]. Therefore, open surgery that ensures anatomical restoration is suggested in patients with severely displaced Gartland type III fractures when closed reduction fails. Posterior open reduction showed worst results, whereas medial and lateral open reduction groups showed good to excellent results [9–11]. At our institution, the medial mini-open approach is applied routinely for open reduction.

in patients with severely displaced Gartland type III fractures when closed reduction fails. Posterior open reduction showed worst results, whereas medial and lateral open reduction groups showed good to excellent results [9–11]. At our institution, the medial mini-open approach is applied routinely for open reduction. Although the skeletons of the older children are similar to adults, they still have growth potential because the epiphyseal plate is open. Kirschner wire was preferred in the fixation of the fracture, but the mechanical strength of the internal fixation needs be higher in older children. Several studies have shown the biomechanical superiority of crossed K-wires [12,13]. At our institution, three crossed K-wires were used for fixation in older children. In this study, we reported the clinical and radiographic long-term effects of children older than 10 years of age with Gartland type III supracondylar humerus fractures who were treated with open reduction by the medial approach, followed by crossed K-wires fixation.

ed K-wires were used for fixation in older children. In this study, we reported the clinical and radiographic long-term effects of children older than 10 years of age with Gartland type III supracondylar humerus fractures who were treated with open reduction by the medial approach, followed by crossed K-wires fixation. Patients and methods Our Institutional Review Board approved the protocol for this project. We retrospectively reviewed the medical records and radiographs of 607 children who underwent open reduction for type III supracondylar humerus fractures at our institution between January 2010 and December 2015. Ninety patients were older than 10 years of age. Patients were included if they fulfilled the following criteria: widely displaced supracondylar humerus fractures (Gartland type III injuries) without vascular injury (Fig. 1), older than 10 years of age with open epiphyseal plates of the distal humerus, with a minimum follow-up of 1 year. Skeletally mature adults and patients with associated bony injuries in the same limb and fractures treated with closed reduction and incomplete data were excluded. A total of five patients could not be traced and their data were excluded. One patient with associated radius fracture in the same limb was excluded. One patient with vascular injury was excluded. Eighty-three patients fulfilled these criteria and there were 61 (73.5%) boys and 22 (26.5%) girls. Seventy-one (85.5%) patients had extension-type fracture and 12 (14.5%) had a flexion-type fracture. The mode of trauma was as follows: fall while playing in 50 (60.2%), fall from height in 23 (27.7%), traffic accidents in seven (8.4%) and blunt trauma in three (3.6%). Associated nerve palsy was radial nerve palsy in two (2.4%), median nerve palsy in two (2.4%) and ulnar nerve palsy in one (1.2%). All of the patients were hospitalized and cross pinning after open reduction by the medial approach was applied.

(27.7%), traffic accidents in seven (8.4%) and blunt trauma in three (3.6%). Associated nerve palsy was radial nerve palsy in two (2.4%), median nerve palsy in two (2.4%) and ulnar nerve palsy in one (1.2%). All of the patients were hospitalized and cross pinning after open reduction by the medial approach was applied. Fig. 1 Anteroposterior and lateral radiographic views showing a Gartland type III supracondylar fracture of the humerus in a 10-year-old girl.

(27.7%), traffic accidents in seven (8.4%) and blunt trauma in three (3.6%). Associated nerve palsy was radial nerve palsy in two (2.4%), median nerve palsy in two (2.4%) and ulnar nerve palsy in one (1.2%). All of the patients were hospitalized and cross pinning after open reduction by the medial approach was applied. Fig. 1 Anteroposterior and lateral radiographic views showing a Gartland type III supracondylar fracture of the humerus in a 10-year-old girl. Surgical technique Surgery was performed under general anaesthesia by a senior surgeon in all 83 patients. With the patient supine, the injured limb was placed on the operating table in abduction and external rotation. A medial incision was made starting from the medial epicondyle and extending proximally for 3–4 cm. The ulnar nerve, often displaced anteriorly in a flexed elbow, was identified and protected. The fracture site was then approached by the intermuscular interstice and the fracture haematoma was drained. The entire anterior and medial bone was well visualized. The elbow was flexed and gentle traction was applied to disengage and visualize the distal fragment. After achieving anatomical reduction, two k-wires (1.5 mm or 2.0 mm in diameter) were inserted percutaneously from lateral to medial. The medial wire (1.5 mm in diameter) was placed percutaneously 0.5–1 cm from the edge of the incision and beginning at the medial epicondyle to engage the lateral cortex. The distance between the crossed k-wires on the fracture plane is more than 1/3 of the width of the humerus on the same plane (Fig. 2). After checking for the capillary refill of the subcutaneous tissue and elbow movements, the wound was irrigated with saline and closed in two planes with a continuous intradermal suture. The wires were bent and cut long to facilitate subsequent removal without anaesthesia. A check radiograph was obtained postoperatively. A long-arm plaster cast with the elbow flexed at 90° was used in all patients for 30 days. Both the parents and the children were instructed on rehabilitative exercises to be continued at home.

were bent and cut long to facilitate subsequent removal without anaesthesia. A check radiograph was obtained postoperatively. A long-arm plaster cast with the elbow flexed at 90° was used in all patients for 30 days. Both the parents and the children were instructed on rehabilitative exercises to be continued at home. Clinical and radiographic examination All patients were followed up for at least 1 year after the initial treatment. Follow-up monitoring included regular clinical and radiographic examination of the patients at 1 and 2 and 12 months after surgery. The cast and the wires were removed at the 4–6-weeks follow-up appointment when osteal callus formed and the fracture line blurred. K-wires would be kept for two more weeks if there was not enough callus around the fracture line. Radiographic evaluation included anteroposterior and lateral radiographs of the injured elbow, which were prepared at each follow-up examination. The healing processes of fracture were confirmed by the radiographs (Fig. 3). The humeroulnar angle was calculated on the anteroposterior radiograph. The humeroulnar angle was defined by the intersection of the midhumeral line with the line drawn from the proximal midpoint of the ulna to the distal midpoint of the ulna in the anteroposterior view on a radiograph with the elbow extended to 0° and the forearm supinated. Fig. 2 Radiography was used for confirmation of the reduction and location of the K-wires. Anteroposterior and lateral views showed satisfactory fracture reduction.

Radiographic evaluation included anteroposterior and lateral radiographs of the injured elbow, which were prepared at each follow-up examination. The healing processes of fracture were confirmed by the radiographs (Fig. 3). The humeroulnar angle was calculated on the anteroposterior radiograph. The humeroulnar angle was defined by the intersection of the midhumeral line with the line drawn from the proximal midpoint of the ulna to the distal midpoint of the ulna in the anteroposterior view on a radiograph with the elbow extended to 0° and the forearm supinated. Fig. 2 Radiography was used for confirmation of the reduction and location of the K-wires. Anteroposterior and lateral views showed satisfactory fracture reduction. Fig. 3 Radiography was used to evaluate the healing of the fracture. Anteroposterior and lateral views 2 months after surgery showed satisfactory fracture restoration. Clinical evaluation included neurovascular examination, measurement of the range of motion of the injured elbow, assessment of the carrying angle, pain and determination of any complications such as infections or iatrogenic nerve injuries. The range of motion of the injured elbow was measured using a manual full-circle goniometer. The clinical results were graded according to Flynn’s criteria, which are based on the carrying angle and the elbow motion [14]. The carrying angle was measured using a full-circle goniometer and compared with that of the contralateral arm.

motion of the injured elbow was measured using a manual full-circle goniometer. The clinical results were graded according to Flynn’s criteria, which are based on the carrying angle and the elbow motion [14]. The carrying angle was measured using a full-circle goniometer and compared with that of the contralateral arm. Statistical analysis Statistical analysis was carried out using the SPSS statistical package program (SPSS 19.0 version; SPSS Inc., Chicago, Illinois, USA). The t-test was performed to compare the injured and uninjured elbows. The null hypothesis was that the mean extension and carrying angle in the injured elbows after fixation would be the same as those in the uninjured elbows (controls). We used a P value less than 0.05 to determine statistical significance of the corresponding variables. Results During the study period, 83 surgically treated paediatric patients with supracondylar humerus fractures fulfilled the criteria for inclusion and were finally enroled in this series (Table 1). Incomplete primary reduction was not encountered in any of our patients. Osteal callus formed and the fracture line blurred at the 30-day follow-up appointment in all patients. Successful fracture healing was achieved 2 months after surgery in all of our patients (100%). The humeroulnar angle averaged 10.1° (range: 1°–17°). No malunions was found in our series (cubitus varus). Table 1 Demographics of the patients

Results During the study period, 83 surgically treated paediatric patients with supracondylar humerus fractures fulfilled the criteria for inclusion and were finally enroled in this series (Table 1). Incomplete primary reduction was not encountered in any of our patients. Osteal callus formed and the fracture line blurred at the 30-day follow-up appointment in all patients. Successful fracture healing was achieved 2 months after surgery in all of our patients (100%). The humeroulnar angle averaged 10.1° (range: 1°–17°). No malunions was found in our series (cubitus varus). Table 1 Demographics of the patients Many scoring systems have been used for elbow disorders. In this study, Flynn’s criteria were used to evaluate the functional and cosmetic results (Table 2). Although 78 (94.0%) patients achieved excellent results and two (2.4%) patients achieved good functional results, three (3.6%) were graded as unsatisfactory because of limited elbow motion. In the remaining three patients classified as having achieved unsatisfactory results, two had an extension deficit of 16° and one had a flexion deficit of 20°. None of the patients had a cubitus varus. Seventy-five (90.4%) patients achieved excellent results, six (7.2%) patients achieved good results and two (2.4%) patients achieved fair cosmetic results. The mean carrying angle measured 9.4° (1°–16°) compared with 10.8° on the contralateral side (5°–16°; Table 3). At the 1-year follow-up examination, none of the patients complained about any relevant pain symptoms. Five (6.0%) peripheral nerve lesions were observed in the first physical examination at admission and resolved spontaneously after an average of 15 weeks (range: 8–20 weeks). There was no case of iatrogenic ulnar nerve injury. Four (4.8%) superficial pin infections were found at follow-up and resolved by appropriate dressing with povidone iodine solution. No antibiotics were administered again.

ination at admission and resolved spontaneously after an average of 15 weeks (range: 8–20 weeks). There was no case of iatrogenic ulnar nerve injury. Four (4.8%) superficial pin infections were found at follow-up and resolved by appropriate dressing with povidone iodine solution. No antibiotics were administered again. Table 2 Flynn’s criteria Table 3 Outcomes according to Flynn’s criteria

ination at admission and resolved spontaneously after an average of 15 weeks (range: 8–20 weeks). There was no case of iatrogenic ulnar nerve injury. Four (4.8%) superficial pin infections were found at follow-up and resolved by appropriate dressing with povidone iodine solution. No antibiotics were administered again. Table 2 Flynn’s criteria Table 3 Outcomes according to Flynn’s criteria Discussion Supracondylar humerus fracture is one of the most common injuries in children, accounting for 50–80% of all paediatric elbow fractures [15]. During the past 5 years, more than 600 patients were treated for supracondylar humerus fracture in our department. About 12% of the patients were children older than 10 years of age. The main goals in paediatric supracondylar fracture treatment are as follows: achievement of normal cosmetic view of elbow, full recovery of elbow movements and protecting the patient from neurovascular complications that may occur. A common complication of paediatric supracondylar fractures is cubitus varus [2]. Biomechanical tests have shown that stress and strain of the medial part of the distal humerus are greater than the lateral part under the axial load. Internal rotation of the distal fragment is the major predisposing factor to varus deformity and is necessary for coronal varus tilt [7]. Stable reduction and fixation are important to prevent displacement of the distal fragment and postoperative deformity [16]. The appropriate strength of fixation was different because the stress of the fracture varies with age and type. Several studies have shown higher incidence of displacement of supracondylar humerus fractures in older children [3]. Satisfactory results can be achieved in Gartland type I and II fractures through closed reduction and immobilization by plaster or percutaneous pinning, which is not feasible in Gartland type III fractures as the fracture is usually associated with extremely swelling in the elbow region. Closed reduction may fail in situations such as delayed presentation and shortage of imaging facility. Soft tissue injuries and stripping of the periosteum resulting from repeated manual reduction were similar to those with open reduction. Percutaneous pinning may be dangerous as surface marking of bony landmarks is difficult. Royce et al. [17] treated 143 children with percutaneous crossed K-wires and described three ulnar secondary nerve palsies.

pping of the periosteum resulting from repeated manual reduction were similar to those with open reduction. Percutaneous pinning may be dangerous as surface marking of bony landmarks is difficult. Royce et al. [17] treated 143 children with percutaneous crossed K-wires and described three ulnar secondary nerve palsies. Therefore, open reduction and internal fixation have been accepted for these widely displaced fractures with severe swelling not allowing acceptable closed reduction and primary neural disruption.

pping of the periosteum resulting from repeated manual reduction were similar to those with open reduction. Percutaneous pinning may be dangerous as surface marking of bony landmarks is difficult. Royce et al. [17] treated 143 children with percutaneous crossed K-wires and described three ulnar secondary nerve palsies. Therefore, open reduction and internal fixation have been accepted for these widely displaced fractures with severe swelling not allowing acceptable closed reduction and primary neural disruption. Four different surgical approaches have been described and every approach has its own advantages [9–11,18]. Anterior incision offers the advantage of a smaller scar and easy access to structures that might be injured between the fractured fragments. Ersan and colleagues reported that all 46 patients achieved satisfactory results on comparison of the anterior and lateral approach in the treatment of extension-type supracondylar humerus fractures in children. However, it is more technically demanding and excessive retraction of the wound can injure the ulnar nerve. The posterior approach is better than the other approaches in manipulation of fracture fragments, but it leads to poor functional results because of soft tissue injuries and fibrous surgical scars [11]. Medial and lateral open reduction approaches lead to similar cosmetic outcomes and functional results. However, Weiland et al. [19] reported a higher incidence of cubitus varus with the use of the lateral approach, which does not allow complete visualization of the medial column communication and tilt. The medial approach provides good visualization, ensuring the restoration of the medial column, and it is a method that induces the least incision scar. At the same time, the influence on the elbow appearance is minimal because the medial approach is hidden.

te visualization of the medial column communication and tilt. The medial approach provides good visualization, ensuring the restoration of the medial column, and it is a method that induces the least incision scar. At the same time, the influence on the elbow appearance is minimal because the medial approach is hidden. In our study, three patients showed poor functional results, whereas none of the patients showed poor cosmetic results. We believe that more follow-up and physical training under the guidance of rehabilitation doctors may be needed to improve the range of motion in the elbow. Several studies have reported that iatrogenic ulnar nerve injuries are encountered in about 6% of patients with supracondylar fractures during percutaneous pinning, including direct penetration or laceration of the nerve or tacking down the nerve sheet in a nonanatomical position [20]. The medial approach allows fracture reduction under visual guidance, which limits the risk of ulnar nerve injury, and has been recommended by several authors [21]. The middle finger of the surgeon touches the broken end of the fracture through the medial incision to assist reduction. Anatomical reduction can always be achieved under the assistance of the finger. In our study, we did not encounter any iatrogenic ulnar nerve palsy. This finding was in agreement with many studies.

. The middle finger of the surgeon touches the broken end of the fracture through the medial incision to assist reduction. Anatomical reduction can always be achieved under the assistance of the finger. In our study, we did not encounter any iatrogenic ulnar nerve palsy. This finding was in agreement with many studies. In the management of Gartland type III fractures, different pin configurations are used. Yaokreh et al. [22] reported that the postoperative secondary displacement rates were 10.4% in a case-series of cross-wire fixation. Flynn et al. [14] described three 72 (4.2%) patients with a cosmetically unsatisfactory result because of loss in the carrying angle. Shim and Lee [23] reported one patient with a cubitus varus deformity in their series of 63 (1.6%) patients treated by cross-fixation with three K-wires. Crossing of the pins in the fracture site is associated with secondary displacement. Skaggs et al. [24] recommended using three diverging lateral epicondylar pins when concerns arose about the stability of the fixation. The internal fixation should be more stable for older children. Several studies have shown that cross pinning provides more stable fixation than lateral diverge or parallel pinning, and reduces the incidence of cubitus varus because of the displacement of the distal part of the fracture [9,10]. In our study, fractures were fixed with three crossed K-wires (two inserted from the lateral side, followed by one from the medial side). Keeping the lateral K-wires apart and the cross points of these three K-wires proximal to the fracture line can lead to more stable fixation. No loss of reduction was found in our series. However, the carrying angle of the injuried elbow reduced 10°–15° compared with the contralateral side in two patients. It may be because the medial cortical compression remained when the lateral cortical continuity was restored. Considering the cosmetic factor, we should compare the injured side with the normal side to ensure that the carrying angle of elbows is similar.

ced 10°–15° compared with the contralateral side in two patients. It may be because the medial cortical compression remained when the lateral cortical continuity was restored. Considering the cosmetic factor, we should compare the injured side with the normal side to ensure that the carrying angle of elbows is similar. In terms of our method of treatment, none of these patients had a significant cubitus varus or iatrogenic ulnar nerve injury. Also, no Volkmann ischaemic contractures or compartment syndromes were observed. All patients were followed up from 12 to 15 months, which may be considered adequate for screening possible complications. One limitation of this study was the lack of a control group for comparison. Conclusion Open reduction by the medial approach and cross-fixation with three kirschner wires for severely displaced Gartland type III fractures is a safe and effective method with a low incidence of complications for older children. Acknowledgements All authors have participated sufficiently in this work concerning conception and design of this study, drafting the article, critical revision for important intellectual content, and final approval. Conflicts of interest There are no conflicts of interest.

Introduction Fractures of the forearm are very common in children and account for more than 30% of all paediatric fractures [1–3]. Angularly deformed forearm fractures are traditionally treated by closed reduction followed by cast immobilisation. Surgical stabilisation is increasingly used as a treatment option, probably due to a relatively high failure rate in the sometimes unpredictable outcome of conservative treatment [4–6]. Redisplacement is the most common complication, especially in primary dislocated forearm fractures (21–40%) [7,8]. Redisplacement or secondary worsening of angulation can be prevented by surgical intervention using percutaneous pinning, intramedullary nailing or plate fixation, which gives maximum stability and the benefit of regaining proper alignment. Fortunately, not all fractures are unstable and require surgical stabilisation since juvenile bone has the unique potency to remodel [9,10]. There is little evidence supporting guidelines on angular acceptance [11]. The uncertainty of predicting fracture stability and the remodelling potential in forearm fractures hinders making a considered decision between conservative and surgical treatment [5,8,12]. Also, there is no convincing literature proving that surgical intervention is superior to conservative treatment in terms of functional outcome [5,8,13].

cting fracture stability and the remodelling potential in forearm fractures hinders making a considered decision between conservative and surgical treatment [5,8,12]. Also, there is no convincing literature proving that surgical intervention is superior to conservative treatment in terms of functional outcome [5,8,13]. The limits of acceptable angular deformations are currently based on scarce retrospective studies, case reports and expert opinions [14–16]. Crawford et al. demonstrated that even completely overriding distal radial fractures have the potential to remodel in one year without reduction [17]. On duration of remodelling, both Friberg et al. 1979 and Jeroense et al. found remodelling speed to be faster in larger angulations [15]. This suggests that deformities can remodel in time and result in a normal functional outcome without experiencing the psychological distress of undergoing a surgical procedure, not to mention exposure to anaesthetic and operative risks. Operative risks should not be underestimated, as earlier studies found a complication rate of 14.6% in patients treated with intramedullary nailing [6]. Although research on fracture remodelling is of great importance in clinical decision-making, to our knowledge no prospective studies have been conducted investigating fracture reangulation in time in conservatively treated paediatric forearm fractures as related to function.

ts treated with intramedullary nailing [6]. Although research on fracture remodelling is of great importance in clinical decision-making, to our knowledge no prospective studies have been conducted investigating fracture reangulation in time in conservatively treated paediatric forearm fractures as related to function. The objective of this prospective study is therefore to first get an initial impression of fracture remodelling and functional outcome in nonreduced paediatric forearm fractures, and second to establish which factors influence remodelling and to determine whether functional outcome is correlated with degree of fracture angulation. Methods Study design and participants This prospective single-centre cohort study was conducted at Isala Clinics in Zwolle, the Netherlands. Children and their parents were verbally informed about the study and also received detailed written information. Informed consent was obtained from the parents and from all children aged ≥12 years only if the child was willing to participate. This study is approved by the local Medical Ethical Committee (CCMO NL12576.075.06). Boys (age <14 years) and girls (age <12 years) with a traumatic angular deformity of the radius, confirmed on postero-anterior and lateral radiographs, were included. Fracture types included comprised isolated radius fractures (plastic deformation or complete fracture) and both-bone forearm fractures.

NL12576.075.06). Boys (age <14 years) and girls (age <12 years) with a traumatic angular deformity of the radius, confirmed on postero-anterior and lateral radiographs, were included. Fracture types included comprised isolated radius fractures (plastic deformation or complete fracture) and both-bone forearm fractures. Exclusion criteria were fully ossified physes of the forearm, manipulated fractures, fracture dislocation, apposition and open fractures. Also excluded were polytrauma patients and patients with a bone disease or pathologic fracture. Maximum acceptable angulations according to age were defined according to the Isala Graphs minus one SD, shown in Table 1 [14]. These graphs are based on the outcome of a meta-analysis of existing literature, combined with the opinions of 18 international experts. Table 1 Maximum acceptable angulations according to age Procedures All fractures were treated with cast immobilisation for 4–6 weeks. On the day of presentation at the hospital (T0), general patient data were collected, including age, gender and hand preference. Patients and their parents were requested to return to the hospital for five follow-up appointments. These sessions were scheduled at 1 week (T1), 4 weeks (T2), 6 weeks (T3), 6 months (T4) and 12 months (T5) postinjury. An optional appointment (T6) was offered when remodelling was delayed.

, gender and hand preference. Patients and their parents were requested to return to the hospital for five follow-up appointments. These sessions were scheduled at 1 week (T1), 4 weeks (T2), 6 weeks (T3), 6 months (T4) and 12 months (T5) postinjury. An optional appointment (T6) was offered when remodelling was delayed. Data collection To determine angular alignment, postero-anterior and lateral radiographs were taken at all follow-up sessions. Degree of angulation was defined as the angle between the central longitudinal intramedullary axis of the proximal and the angulated distal fragment as previously described by Hansen et al. (1976) [18] Measurements were taken by two independent observers who were not involved in the treatment (J.J.W.P. and B.B.). The largest angulation at T0 (on the postero-anterior or lateral radiograph) was further observed during follow-up. Additionally, grip strength and passive range of motion of the wrist were tested for both hands at each follow-up appointment, with the exception of T0. Tests were not performed at T0 due to the cast immobilisation. Grip strength measurements were taken using a Jamar hydraulic hand dynamometer (Sammons Preston Rolyan, Chicago). Grip strength was tested twice on both sides and the mean score of the two attempts for each side was used in the analyses. Passive range of motion was measured using a goniometer, and included flexion and extension of the elbow, pronation and supination of the forearm, and palmar and dorsal flexion and ulnar and radial deviation of the wrist.

twice on both sides and the mean score of the two attempts for each side was used in the analyses. Passive range of motion was measured using a goniometer, and included flexion and extension of the elbow, pronation and supination of the forearm, and palmar and dorsal flexion and ulnar and radial deviation of the wrist. Statistical analyses All statistical analyses were conducted using SPSS (version 24.0, SPSS Inc., Chicago). Descriptive statistics were used to describe the main characteristics of the research population and functional outcome parameters. The mean angular deformity as determined by both observers was used in the analyses, as interrater reliability appeared to be excellent (intraclass correlation coefficient 0.98). The Wilcoxon signed-rank test was used to compare grip strength and range of motion of the affected and unaffected hands. A multilevel design was applied, which implies that the follow-up appointments were nested under patients. A multiple regression analysis was performed with fracture angulation as dependent variable. The following factors were tested for association with the above-mentioned variable: time post-injury, dominant arm fractured, type of fracture (plastic deformation or complete fracture) and involvement of the radius or both the radius and ulna. An unconditional growth model will be presented with fracture angulation as dependent variable and time and function tests as independent variables. Results were accepted as significant if P < 0.05.

d, type of fracture (plastic deformation or complete fracture) and involvement of the radius or both the radius and ulna. An unconditional growth model will be presented with fracture angulation as dependent variable and time and function tests as independent variables. Results were accepted as significant if P < 0.05. Results Patient characteristics A total of 27 children were enrolled in this study. One child did not show up at the follow-up appointments and was therefore excluded. The final study population comprised 26 children (13 boys), ranging from ages 3.3 to 12.6. Mean age at the time of injury was 9 years (boys: 9.1; girls: 8.9). Of all children, 88.5% were right-hand dominant and 17 fractures (65.4%) affected the nondominant side. In 38.5% the fracture concerned a plastic deformation; 61.5% had a complete fracture (both cortices). This was equally distributed between both sexes. Most boys (61.5%) sustained a both-bone fracture, whereas most girls (69.2%) sustained a solitary radius fracture. All fractures were distally located except in two cases with a midshaft both-bone fracture. All fractures were conservatively treated with cast immobilisation. Mean immobilisation time was 28 days (SD 5.3). The main characteristics of the study population are shown in Table 2. Table 2 Characteristics of the study population

Results Patient characteristics A total of 27 children were enrolled in this study. One child did not show up at the follow-up appointments and was therefore excluded. The final study population comprised 26 children (13 boys), ranging from ages 3.3 to 12.6. Mean age at the time of injury was 9 years (boys: 9.1; girls: 8.9). Of all children, 88.5% were right-hand dominant and 17 fractures (65.4%) affected the nondominant side. In 38.5% the fracture concerned a plastic deformation; 61.5% had a complete fracture (both cortices). This was equally distributed between both sexes. Most boys (61.5%) sustained a both-bone fracture, whereas most girls (69.2%) sustained a solitary radius fracture. All fractures were distally located except in two cases with a midshaft both-bone fracture. All fractures were conservatively treated with cast immobilisation. Mean immobilisation time was 28 days (SD 5.3). The main characteristics of the study population are shown in Table 2. Table 2 Characteristics of the study population Radiographic outcome An overview of radiographic outcome is shown in Table 3. All maximum angulations occurred in the sagittal plane. Dorsal angulation occurred in 65.4% of cases. Mean angulation was 11.7° (5.0–18.0) at the day of presentation, 11.8° (4.0–22.5) after 1 week and 12.8° (4.0–22.0) after 4 weeks. Six months after sustaining the fracture, the mean angulation diminished to 6.3° (1.0–10.5) and after 1 year to 3.6° (0.0–13.0), with fracture angulation amounting to less than 5° in 75% of cases. The distribution of fracture angulation is shown in Fig. 1. One outlier remained as a residual angulation of 13° one year postinjury. This concerned a 12-year-old boy with a midshaft both-bone fracture. Because of the remaining angulation, a control radiograph was taken 2.9 years after fracture sustainment. Angulation remained at 11°. Mean angulation and distribution for each follow-up moment is plotted in Fig. 2.

gulation of 13° one year postinjury. This concerned a 12-year-old boy with a midshaft both-bone fracture. Because of the remaining angulation, a control radiograph was taken 2.9 years after fracture sustainment. Angulation remained at 11°. Mean angulation and distribution for each follow-up moment is plotted in Fig. 2. Table 3 Outcome of fracture angulation Fig. 1 Fracture angulation distribution in % for each follow-up appointment. Fig. 2 Mean dorsal angulation (°) and distribution (SD) plotted in time. The line represents the mean dorsal angulation. Functional outcome Grip strength Grip strength is significantly diminished in the affected hand compared to the unaffected hand up to 6 months postinjury (T4). The results show that grip strength is strongly diminished at T1, T2 and T3, and less diminished but still significant at T4. After one year follow-up, grip strength measurements showed no significant difference between the affected and unaffected arm. When describing grip strength of the affected side compared to the unaffected side (%), results show a mean grip strength of 97% at both T4 and T5 (T4: SD 17.6, T5: SD 14.1). An overview of recovery of grip strength is shown in Table 4 and Fig. 3. Table 4 Grip strength of affected hand vs. unaffected hand Fig. 3 Mean grip strength of the affected arm presented as percentage of the unaffected arm.

Functional outcome Grip strength Grip strength is significantly diminished in the affected hand compared to the unaffected hand up to 6 months postinjury (T4). The results show that grip strength is strongly diminished at T1, T2 and T3, and less diminished but still significant at T4. After one year follow-up, grip strength measurements showed no significant difference between the affected and unaffected arm. When describing grip strength of the affected side compared to the unaffected side (%), results show a mean grip strength of 97% at both T4 and T5 (T4: SD 17.6, T5: SD 14.1). An overview of recovery of grip strength is shown in Table 4 and Fig. 3. Table 4 Grip strength of affected hand vs. unaffected hand Fig. 3 Mean grip strength of the affected arm presented as percentage of the unaffected arm. Range of motion At T2 the affected hand scored significantly lower in all mobility tests, except for extension of the elbow. At T3 the affected hand scored significantly lower in all mobility tests, except for radial deviation. Six months postinjury (T4), only pronation (P ≤ 0.01) and supination (P = 0.03) were significantly diminished in the affected arm. Range of motion after one year follow-up (T5) showed no statistically significant differences in elbow and wrist motion of the affected arm compared to the unaffected arm (see Table 5). Maximum loss of range of motion at T5 was found to be 10° in radial deviation and pronation. Table 5 Range of motion after 1-year follow-up

Range of motion after one year follow-up (T5) showed no statistically significant differences in elbow and wrist motion of the affected arm compared to the unaffected arm (see Table 5). Maximum loss of range of motion at T5 was found to be 10° in radial deviation and pronation. Table 5 Range of motion after 1-year follow-up Factors affecting remodelling A multiple regression analysis with fracture angulation as dependent variable shows that fracture angulation significantly diminishes in time (adjusted coefficient = −0.03, P ≤ 0.01). Greenstick fractures show significantly faster remodelling than full-thickness fractures (adjusted coefficient = −3.04, P = 0.0145). An affected dominant or non-dominant hand, as well as suffering from a solitary radius fracture or both-bone fracture, is not of significant influence on fracture angulation. Fracture angulation and function Using unconditional growth model analyses, grip strength was found to be significantly influenced by fracture angulation (coefficient = −1.52, P = 0.0223). No association was found between fracture angulation and any range of motion tests. Discussion The current study shows a first impression of the bone remodelling capacity in nonreduced paediatric forearm fractures, thereby evaluating functional outcome in time. Factors that influence fracture angulation were also determined. The rationale was the lack of clear guidance from the literature for definite acceptable angular deformations and functional restoration in time.

modelling capacity in nonreduced paediatric forearm fractures, thereby evaluating functional outcome in time. Factors that influence fracture angulation were also determined. The rationale was the lack of clear guidance from the literature for definite acceptable angular deformations and functional restoration in time. After one year, the mean fracture angulation of 12° measured at initial presentation was reduced to a mean residual angulation of 4°. At this point in time, no significant differences between the affected and the nonaffected hand were found for either grip strength or range of motion. This suggests that a residual angulation of 4° is of no functional concern. Conservative treatment without reduction could therefore be a good treatment option in angulated forearm fractures. There is a worldwide tendency toward a more aggressive approach in the treatment of the described angular deformities, even without thoroughly weighing noninvasive treatment modalities. Using the Isala Graphs as a safe inclusion, we attempted to obtain more insight into functional outcome in nonreduced angulated fractures. Functional outcome is often overlooked while in daily practice fracture consolidation often equals the end of treatment.

ithout thoroughly weighing noninvasive treatment modalities. Using the Isala Graphs as a safe inclusion, we attempted to obtain more insight into functional outcome in nonreduced angulated fractures. Functional outcome is often overlooked while in daily practice fracture consolidation often equals the end of treatment. Literature on fracture remodelling in paediatric forearm fractures is scarce, even more so in relation to functional outcome [19]. Crawford et al. performed a retrospective case series amongst 54 children with conservatively treated overriding distal radius fractures [17]. Angulation improved from 4.0 to 2.2° at final follow-up (one-year after fracture) with no functional limitations. Functional outcome during follow up and final scores were, however, not specified and could therefore not be compared to the recent study. Also, their study population consisted of completely displaced fractures and were excluded in our study. In a retrospectively studied population of 33 children with malunited distal radius fractures, Jeroense et al. found a mean residual angulation of 8° after a mean follow-up of 9 months, compared to 4° residual angulation after 12 months in our study population. However, mean angulation at moment of presentation was larger in their population (23°) than in ours (12°) [15]. The study of Van der Sluijs et al. (2016) merged data of two studies (including Jeroense et al. 2015), and included 63 children with a mean angulation at initial trauma of 25°, which remodelled to a mean residual angulation of 6.7° after a mean of 22 months follow-up [16]. Neither of these studies took functional outcome into consideration though.

t al. (2016) merged data of two studies (including Jeroense et al. 2015), and included 63 children with a mean angulation at initial trauma of 25°, which remodelled to a mean residual angulation of 6.7° after a mean of 22 months follow-up [16]. Neither of these studies took functional outcome into consideration though. As mentioned in the Results section, one case maintained a residual angulation of 11°. Stagnation of remodelling in this case could be partially explained by fracture location and age. More proximally located fractures of the radius and ulna are known to have a high probability of residual angulation and pronation loss [20]. Johari et al. described how midshaft forearm fractures in children older than age 10 have a less favourable prognosis in terms of remodelling [19]. Despite the residual angulation, grip strength and range of motion were found to be near-normal, with all scores being equal to the unaffected side except for pronation and grip strength. These scores were both 90% of the unaffected hand. The minimum loss of function despite the residual malalignment of 11° could be explained by the extent of malalignment. Colaris et al. (2014) found a significant loss of pronation (<50°) more than 6 months post-trauma in 31.9% of cases with an angular malunion of 11–15° [20]. Earlier cadaveric studies with artificially created deformities of the forearm bones revealed that angular malalignment of 10° or less will not limit forearm rotation anatomically, while loss of pronation and supination can be expected when residual angles of 20° or more are measured [21,22].

r malunion of 11–15° [20]. Earlier cadaveric studies with artificially created deformities of the forearm bones revealed that angular malalignment of 10° or less will not limit forearm rotation anatomically, while loss of pronation and supination can be expected when residual angles of 20° or more are measured [21,22]. As expected, the radiographs show a reduction of angulation over time. Interestingly, in some cases, angulation seems to increase in the first period before a decrease sets in. This phenomenon is not previously described in studies on nonreduced forearm fractures. However, Colaris et al. described an angulation increase in forearm fractures treated by reduction, in the period between reduction and cast removal [7]. In his study, as in ours, remodelling was seen in the period between cast removal and final examination. Previous studies have shown that fractures with any bayonet apposition are prone to lose reduction, which could probably explain the primary worsening of angulation [23,24].

d between reduction and cast removal [7]. In his study, as in ours, remodelling was seen in the period between cast removal and final examination. Previous studies have shown that fractures with any bayonet apposition are prone to lose reduction, which could probably explain the primary worsening of angulation [23,24]. It would be reasonable to assume that after correction of angular deformity in time, recovery of function would follow. This study found an excellent functional outcome after one year. No significant differences in elbow or wrist range of motion were observed between the affected and the unaffected arm. Pronation and supination took the longest to recover since the scores on these parameters where both still significantly diminished up to 6 months after fracture. This observation is in line with previous literature, where limitations in pronation and supination were most frequently seen in overall mobility of the wrist after sustaining a forearm fracture [25,26]. The potency of angular correction in juvenile bone depends on redirection of the epiphyseal growth plate and remodelling at the fracture site [27,28]. An interesting thought would be that remodelling is being promoted by function. Factors supporting this can be derived from for example, Wolff’s law; malalignment in plane of movement is advantageous and rotational deformities in a fracture do not realign. Redistribution of growth in the physis still remains hard to prove [27,29].

ng thought would be that remodelling is being promoted by function. Factors supporting this can be derived from for example, Wolff’s law; malalignment in plane of movement is advantageous and rotational deformities in a fracture do not realign. Redistribution of growth in the physis still remains hard to prove [27,29]. To our knowledge, this study is the first to prospectively investigate fracture remodelling in paediatric nonreduced angulated forearm fractures for functional outcome at fixed follow-up moments. Since assessment took place several times during one year, this study provides good insight into the progression of remodelling as well as recovery of function over time. The most important limitation of this study is the relatively small study population. This makes the data less reliable to adequately differentiate between subgroups (e.g. hand dominance, sex and fracture type). Second, the range of fracture angulation at the moment of presentation was large and relatively moderate because of the inclusion criteria. More subjects are needed to adequately observe the difference in fracture remodelling, based on severity of angulation at the time of presentation. Lastly, we had to deal with missing data. Not all participants came to all the follow-up appointments. For future studies, the recommendation would be to schedule less follow-up appointments at stricter times to improve attendance.

ce in fracture remodelling, based on severity of angulation at the time of presentation. Lastly, we had to deal with missing data. Not all participants came to all the follow-up appointments. For future studies, the recommendation would be to schedule less follow-up appointments at stricter times to improve attendance. Conclusion This study shows that nonreduced angulated paediatric forearm fractures have the potential to remodel in time, and show good radiographic and functional outcome with respect to grip strength and range of motion after one year. Concerning functional outcome, pronation and grip strength take the longest to recover, with grip strength being strongly associated with fracture alignment. Acknowledgements The help of R.E. Steward, Methodologist at the University Medical Center Groningen, is greatly appreciated with conducting the analyses. Conflicts of interest There are no conflicts of interest.