Journal of Hand Surgery (European Volume)

Acute scaphoid fractures are common wrist injuries that continue to elicit debate from surgeons regarding the most appropriate diagnostic and management algorithms. This review will examine the current literature and trends, in an attempt to provide the reader with an evidence-based discussion regarding current controversies of interest to clinicians. In addition, an attempt will be made to provide recommendations for the best treatment practices for acute scaphoid fractures.

Treatment of scaphoid fractures has been focused mainly on achieving union, with less attention to restoring normal scaphoid shape and orientation. Malalignment of one carpal bone will disrupt the kinetics of the entire wrist. The dorsal intercalated segment instability deformity associated with scaphoid waist nonunion is a nondissociative form of carpal instability. It has to be reduced in the treatment of scaphoid waist nonunions to avoid kinetic problems that will lead to arthritic changes. Computerized tomography scanning has become indispensable to visualize the humpback deformity clearly. Different techniques may be used to restore the normal anatomy of scaphoid, from non-vascularized graft to arthroscopic bone grafting, and also the option of vascularized bone grafting.

We present the surgical outcomes of non-vascularized bone grafting taken from the iliac crest in 24 patients with scaphoid nonunion and avascular necrosis. The Fisk-Fernandez technique was used in 11 patients, and cancellous bone grafting was used in 13 patients. Bony union was achieved in 22 of the 24 patients. Non-vascularized iliac bone grafting can be used for the surgical management of scaphoid nonunion with avascular necrosis. Although revascularization of the proximal fragment after surgery was not evaluated, bony union was confirmed in nearly all patients.

Level of evidence: IV

We followed 35 consecutive patients with scaphoid nonunions in a prospective longitudinal registry. All nonunions were treated with curettage, non-vascularized autogenous grafting and headless screw fixation. Preoperative magnetic resonance imaging, intraoperative bleeding points and histopathological analysis of cancellous bone in the proximal pole were recorded as measures of viability. Healing was categorized as ≥50% bony bridging on computed tomographic images in the plane of the scaphoid. Nine of 23 proximal poles demonstrated ischaemia on magnetic resonance imaging but none were interpreted as infarcted. Twenty-eight of 33 were found to have impaired vascularity as assessed by intraoperative bleeding. Fourteen of 32 demonstrated ≥50% trabecular necrosis and four of 33 demonstrated ≥50% tissue necrosis on histopathological analysis. Thirty of 33 demonstrated focal or robust remodelling activity. Despite pathological evidence of impaired vascularity in over half of the patients, 33 of the 35 scaphoids had healed by 12 weeks. We conclude that proximal pole infarction is decidedly rare and that vascularized bone grafting is seldom required.

Level of evidence: IV

The reported results for the treatment of scaphoid nonunions with non-vascularized grafts are based on observational studies with a high variability in union rates from 38%–100% and time to union from 6–18 weeks, and there is also a lack of a standard classification. Meta-analyses and systematic reviews have been presented to better support conclusions from large pools of data (604 to 5246 cases), but their interpretation remains limited because of lack of uniformity in the studies analysed. Several authors have presented results from the Zaidemburg flap, with highly variable outcomes despite using the same technique; union rates have ranged from 27%–100% achieved between 6.5–19 weeks. Technical details that may be responsible for the discrepancies are discussed. There is a need for a validated prognostic classification system for scaphoid nonunions that can allow comparisons between outcome studies.

Vascularized osteochondral flaps have been described for use in reconstruction of the wrist, providing the benefits of osteochondral grafts in addition to the benefit of osseous and subchondral perfusion via the microvascular pedicle. Various harvest sites have been described including the medial and lateral femoral trochlea and the proximal third metatarsal. The reconstructed surfaces described include the cartilage surfaces of the scaphoid, lunate, capitate and radius. The ability to transfer analogous osteochondral surfaces poses the possibility for new motion-preserving alternatives for difficult articular problems previously treated with salvage procedures. A description of the procedures and reported outcomes are provided, as well as a discussion of the role of these procedures in the future of wrist surgery.

Treating scaphoid nonunions presents difficulties particularly when there is bone loss, significant humpback deformity or avascular necrosis. We describe a new type of fixation with a volar scaphoid plate that adds to the methods of internal fixation that are available for the treatment of recalcitrant scaphoid nonunions. We will also discuss ‘lessons learned’ from a cases series. The case series includes 20 consecutive patients treated with volar buttress plating and a pedicled vascularized bone graft from the ipsilateral volar distal radius. There was clinical and radiographic evidence of union in 18 of 20 patients, 13 of which were verified by computed tomographic scan. The range of motion was improved in all patients post-operatively. Four patients with radiographic union experienced intermittent clicking with maximal wrist flexion, believed to be due to the impingement of the plate on the volar aspect of the radioscaphoid articulation and underwent removal at approximately 1 year after the index procedure. Volar scaphoid plating is a useful alternative to headless scaphoid screw fixation in the treatment of unstable scaphoid waist fractures and nonunions.

Level of evidence: IV

We present 20 patients with established proximal pole scaphoid nonunions treated with curettage and cancellous autograft from the distal radius and screw fixation. Fractures with significant proximal pole fragmentation were excluded. Patients were treated at a mean of 26 weeks after injury (range 12–72). Union occurred in 18 of 20 patients (90%) based on computed tomographic imaging. The two nonunions that did not heal were treated with repeat curettage and debridement and iliac crest bone grafting without revision of fixation. Union was achieved in both at a mean of 11 weeks after the revision procedures. Our findings suggest that non-vascularized cancellous autograft and antegrade fixation is a useful option for the treatment of proximal pole scaphoid nonunions.

Level of evidence: IV

The purpose of this study was to measure the radiographic parameters of proximal pole scaphoid fractures, and calculate the ideal starting points and trajectories for antegrade screw insertion. Computed tomography scans of 19 consecutive patients with proximal pole fractures were studied using open source digital imaging and communications in medicine (DICOM) imaging measurement software. For scaphoid sagittal measurements, fracture inclination was measured with respect to the scaphoid axis. The ideal starting point for a screw in the proximal pole fragment was then identified on the scaphoid sagittal image that demonstrated the largest dimensions of the proximal pole, and hence the greatest screw thread purchase. Measurements were then taken for a standard screw trajectory in the axis of the scaphoid, and a trajectory that was perpendicular to the fracture line. The fracture inclination in the scaphoid sagittal plane was 25 (SD10) °, lying from proximal palmar to dorsal distal. The fracture inclination in the coronal plane was 9 (SD16) °, angling distal radial to proximal ulnar with reference to the coronal axis of the scaphoid. Using an ideal starting point that maximized the thread purchase in the proximal pole, we measured a maximum screw length of 20 (SD 2) mm when using a screw trajectory that was perpendicular to the fracture line. This was quite different from the same measurements taken in a trajectory in the axis of the scaphoid. We also identified a mean distance of approximately 10 mm from the dorsal fracture line to the ideal starting point. A precise understanding of this anatomy is critical when treating proximal pole scaphoid fractures surgically.

Radial styloidectomy along with scaphoid nonunion reconstructive surgery is considered an acceptable surgical treatment for stage 1 scaphoid nonunion advanced collapse. The extent of radioscaphoid joint wear that defines scaphoid nonunion advanced collapse stage 1 is unclear. An appropriate radial styloidectomy is limited to 6 mm styloid removal to prevent carpal instability. The aim of this study was to determine whether nonunion fracture location can be a guiding factor in determining whether the scaphoid bone can be retained with scaphoid nonunion reconstructive surgery and radial styloidectomy. We reviewed wrist radiographs and magnetic resonance imaging studies of 21 patients with a diagnosis of scaphoid nonunion advanced collapse stage 1, and quantified chondral wear as the length of a line (in mm) along the radioscaphoid joint traced from the tip of the radial styloid to the most ulnar position along the radioscaphoid joint. We concluded that scaphoid nonunion reconstructive surgery with radial styloidectomy is only appropriate for nonunions in the distal or middle third of the scaphoid.

Level of evidence: III

The aim of this study was to analyse lunate fractures and any associated osseo-ligamentous injuries. A systematic review identified 34 cases. We identified carpal instabilities at the radiocarpal and midcarpal joints in volar and dorsal directions. Radiocarpal instabilities (10/34) were usually dorsoradial (8/10), with a transverse lunate fracture, best seen on a coronal image. Midcarpal instabilities (24/34) were usually volar (14/18), with a volar lunate shear fracture, best seen on a sagittal image. Instabilities were sub-classified into non-displaced, subluxated and dislocated. Associated fractures of the scaphoid and the radial and ulnar styloid processes were common. Lunate fractures without subluxation or dislocation had good outcomes with cast immobilization or fixation of associated fractures. Lunate fracture-subluxations are unstable injuries that are best managed with fixation of the carpal fractures. Lunate fracture-dislocations are complex injuries, requiring stabilization of the lunate, associated fractures and ligament injuries; complications are common and acute or delayed salvage procedures may be required.

We investigated the effect of an interfragmentary gap on the final compression force using the Acutrak 2 Mini headless compression screw (length 26 mm) (Acumed, Hillsboro, OR, USA). Two blocks of solid rigid polyurethane foam in a custom jig were separated by spacers of varying thickness (1.0, 1.5, 2.0 and 2.5 mm) to simulate an interfragmentary gap. The spacers were removed before full insertion of the screw and the compression force was measured when the screw was buried 2 mm below the surface of the upper block. Gaps of 1.5 mm and 2.0 mm resulted in significantly decreased compression forces, whereas there was no significant decrease in compression force with a gap of 1 mm. An interfragmentary gap of 2.5 mm did not result in any contact between blocks. We conclude that an increased interfragmentary gap leads to decreased compression force with this screw, which may have implications on fracture healing.