We retrospectively reviewed 17 patients with symptomatic diaphyseal malunions of the forearm resulting from pediatric forearm fractures treated by corrective osteotomy. Patients were grouped into three categories according to their main preoperative complaints. We assessed preoperative and postoperative ROM and radiographs, and strength, symptoms, and the occurrence of complications. Statistical analysis was used to assess how the type of preoperative complaint affected ROM and functional improvements after osteotomies.
This series included all patients consecutively operated on by the two senior authors (LN, CED) between 1994 and 2001 in the Department of Orthopaedic Surgery at the University of Zurich and the Division of Hand Surgery at the University of Bern. We based the indication for the osteotomy on stiffness of forearm ROM in pronation-supination or the occurrence of painful snapping at the DRUJ during pronation and supination that impaired activities of daily living. The patient records and radiographs were reviewed at last followup by one independent observer (LJ) who was blinded to the results. The patients’ mean age at the time of surgery was 20.6 years (range, 12.8–41.6 years). There were six female patients and 11 male patients (Table ). We treated nine right and eight left extremities. The dominant arm was affected in six patients and the nondominant arm in 11 patients. They were no Monteggia or Galeazzi fractures based on the normal anatomy of the humeroradial and the distal radioulnar joints assessed on the initial radiographs. The treatment for a forearm fracture occurred at a mean age of 12.8 years (range, 7.1–19.2 years); the mean delay from fracture to corrective osteotomy was 7.8 years (range, 0.4–30.2 years); and the minimum followup was 6 months (mean, 3.7 years; range, 6 months–9.9 years).
Three groups of patients were defined according to the main clinical problem (Table ). Six patients (gender ratio, 1:5 male:female; mean age at fracture, 12.7 years; mean age at osteotomy, 24.4 years) had predominant loss of pronation (Group 1), four (gender ratio, 4:0 male:female; mean age at fracture, 13.0 years; mean age at osteotomy, 17.3 years) had predominant loss of supination (Group 2), and seven (gender ratio, 6:1 male:female; mean age at fracture, 12.8 years; mean age at osteotomy, 19.3 years) had a painful DRUJ but no major pronation-supination impairment in comparison to the healthy side (Group 3).
The level of the deformity was defined as percent of the entire length of the bone (Fig. ). Seventy-two percent of malunions were located in the middle third of the radius and 90% in the middle third of the ulna (Fig. ). In Group 1, all malunions were located in the proximal two-thirds of the radius and the ulna with both forearm bones always involved with angular deformities (Fig. ). Four patients in this group also had axial malunions of 30° or greater affecting one of the forearm bones. In Group 2, all but one patient (Patient 7) had combined angular malunions of the radius and the ulna. Only one (Patient 5) had the malunion located in the distal third of the radius. No patient in this group had a rotational malunion of the radius of 30° or greater, and only one patient (Patient 8) had a combined rotational malunion of the ulna equal to 25°. Conversely, all but one patient (Patient 11) in Group 3 had an isolated malunion of the radius. All patients in this group had malunions of the radius located in the distal half of the bone, six of seven were in the distal third.
The levels of malunions of the radius and the ulna are shown. The number of rectangles indicates the number of osteotomies that were performed at this level of the radius and the ulna.
Preoperative (A) anteroposterior and (B) lateral and postoperative (C) anteroposterior and (D) lateral radiographs show the forearm with malunions of the radius and the ulna corrected with osteotomies of both forearm bones (Patient 8).
For preoperative planning, the opposite healthy side served as a template because the correctional osteotomy intended to reproduce the osseous geometry of the normal side. Therefore, we obtained plain radiographs of both forearms in full length. For anteroposterior and lateral projections of the radius, the distal joint surface was considered, whereas for the ulna, the humeroulnar joint was used as assessed under the image intensifier. The contours of the healthy and deformed bones in both projections were drawn on separate sheets of tracing paper. By simple superposition, we determined the location of maximal deformity and angular deformity in both planes (Fig. ). From these projections, the true angle of deformity (δ), corresponding to the maximal angulation, and its orientation in space (β) were calculated using established tables [29
]. If the anatomic relationship of the radial styloid and the bicipital tuberosity and that of the ulnar styloid and the coronoid process were different on both sides, we suspected an axial malunion. The level of the axial malunion was determined as the level of the initial fracture as assessed in percentages of the entire length of the bone or was considered at the same level as the angular deformity if both deformities were present. The exact amount of radial and ulnar torsion was defined on comparative MR images of both forearms in nine patients (Table ) [4
]. Side difference in the torsion profile of the radius exceeding 30° or 20° in the ulna was a reason to correct the axial malunion, because side differences greater than these limits are considered nonphysiologic [10
]. The site of the osteotomy determined the type of exposure. We marked the position of the planned osteotomy with a Kirschner wire placed under fluoroscopic control. Two Kirschner wires marked the frontal plane in both fragments using the distal radius and the humeroulnar joint as landmarks for the radius and the ulna, respectively (Fig. ). At the planned site, a wedge of bone usually was excised (incorporating the true angle of deformity = δ) in the proper orientation (in the plane of the true deformity = β) allowing for a closing wedge osteotomy, which was instrumented with a 3.5-mm compression plate. Opening wedge osteotomy with interposition of a structural graft was performed when facing relative shortening of the radius to produce ± 1 mm ulnar variance. In case of combined angular and axial malunions, a transverse osteotomy was performed and derotation of the forearm bone was performed using bone clamps to stabilize the bone extremities with the plate. Intraoperative orthogonal radiographs were obtained and the angular deformity was further corrected with attention paid to restoration of the radial bow [34
]. After completion of the correction, the plate was definitively fixed to the bone with screws. Intraoperatively, we checked correctness of spatial reconstruction by standard biplanar radiographs. If reconstruction of the true anatomic shape of the forearm bones did not result in free motion, we identified the interosseous membrane and released it from its ulnar insertion, except in Patient 7, in whom the release was performed from radial insertion. In some cases, reduction of the osteotomy proved difficult or impossible because of the overly tense, contracted interosseous membrane. In these cases, we split the interosseous membrane from ulnar insertion to permit reduction.
Fig. 3A–F A radius malunion consists of an angulation at the middle third of the bone in the radial-dorsal to ulnar-volar plane. The orientation of the deformity in space and the value of the maximal angular deformity, termed true angle of deformity, are assessed (more ...)
Preoperative and postoperative radiologic assessments
We performed 17 osteotomies (one double osteotomy) of the radius, 10 osteotomies of the ulna, two interpositional corticocancellous bone grafts in open-wedge osteotomy of the radius, and eight releases of the interosseous membrane. The forearm was immobilized with a resting splint for 6 weeks. All patients started forearm active motion exercises within the first 2 postoperative weeks. Dynamic stretching splints, passive motion, and strengthening were started after 6 weeks. Overall, six patients had hardware removal before the last followup.
We measured forearm pronation-supination with a gravity goniometer and the value was indicated according to the neutral-null method [40
]. Stability of the distal radioulnar joint was assessed clinically in neutral rotation by manually stressing the joint palmarly and dorsally [22
]. We clinically assessed ulnar impaction syndrome that should have resulted from inadequate shortening of the radius or longitudinal instability of the forearm after splitting the interosseous membrane with a compression test of the ulnocarpal joint and finger palpation of the triangular fibrocartilage complex (TFCC) [21
]. A Jamar dynamometer (Jamar Hand Dynamometer; Sammons Preston Inc, Bollingbrook, IL) at setting II was used to measure grip strength using the average of three measurements. Percentage of strength between sides, termed relative grip strength, was calculated. The first part of the Disabilities of the Arm, Shoulder and Hand (DASH; www.dash.iwh.on.ca) score was administered as a self-report questionnaire to measure function and symptoms at the last followup [17
]. It contains 23 questions to assess impairment in activities of daily living with a scale ranging from 1 (not limited at all) to 5 (unable), six questions to assess pain with a scale ranging from 1 (none) to 5 (extreme), and one question about influence of pain on sleep with a scale ranging from 1 (no difficulty) to 5 (so much difficulty that I cannot sleep). We then normalized the score (observed score-30/1.2) to obtain a linear scale ranging from 0 to 100. The overall scaling direction indicated the higher the score, the worse the functional outcome. We obtained full-length conventional radiographs of both forearms, anteroposterior and lateral in neutral rotation and with the elbow flexed at 90°, and any angular residual deformity was measured in the same way as preoperatively. Residual axial malunions were determined systematically at the last followup with fluoroscopy coupled with goniometry [10
Measurements were recorded on a Microsoft® Office Excel® 2003 data sheet (Microsoft Corp, Walisellen, Switzerland). Mean values and standard deviations were calculated. We used a paired t-test to compare preoperative and postoperative ROM, one-way analysis of variance with Tukey’s honestly significant difference post hoc test to compare gain in ROM among groups of patients, Mann-Whitney U test (release versus nonrelease of the interosseous membrane), or Kruskal-Wallis one-way analysis of variance by ranks to compare age at fracture and osteotomy, grip strength, and DASH scores among groups of patients using StatView 5.01 (SAS Institute, Cary, NC). Significance was set at p < 0.05.