▪ Indications

The indications for percutaneous scaphoid screw insertion include nondisplaced and minimally displaced scaphoid waist fractures that are reducible with percutaneous methods. Distal pole fractures have a better prognosis and are often successfully treated with casting, but screw fixation is indicated for distal pole nonunions. Undisplaced fibrous scaphoid waist nonunions without evidence of avascular necrosis are also suitable candidates provided there are no radiocarpal or midcarpal degenerative changes. Undisplaced proximal pole fractures without avascular necrosis are amenable to this technique provided that the proximal fragment is large enough to allow capture by at least 5 mm of screw threads.1

▪ Contraindications

Severely displaced fractures that cannot be reduced by closed means are not suited to this approach. Partial or complete avascular necrosis of the scaphoid is a relative contraindication to percutaneous screw fixation, although successful cases have been reported.2 A small proximal pole fragment does not allow adequate screw purchase. Nonunions with a humpback deformity and secondary dorsal intercalated segmental instability (DISI) patterns require an open volar wedge graft. The presence of significant radiocarpal and/or midcarpal degenerative changes also mandate an alternate approach.

▪ Preoperative Imaging

The fracture position is assessed with an anteroposterior (AP), lateral, and 45 degree semipronated view, although a computed tomographic (CT) scan is helpful in difficult cases ( Fig. 13.1 ). It is my preference to perform a magnetic resonance imaging (MRI) in all scaphoid nonunions preoperatively to rule out avascular necrosis.

▪ Screw Placement

As a general rule, volar implantation is ideal for distal pole and scaphoid waist fractures. Volar implantation requires eccentric screw placement through the distal pole because one has to ream through the trapezium to gain access to the central scaphoid axis. Despite this oblique distal starting point, the screw can still gain access to the central canal at the level of the waist and proximal pole.3 Screw fixation of the scaphoid through a volar approach is hindered by the trapezium, which prevents a straight-line approach to the central axis of the scaphoid. There is the added risk of screw cutout through the concave volar surface of the scaphoid or through the dorsoulnar aspect of the proximal pole. Furthermore, an overly long screw will impinge upon the scaphotrapezial joint and lead to osteoarthritis ( Fig. 13.2 ).

Guide wire insertion is integral to each procedure since screws that are placed along the central axis provide greater rigidity than eccentrically placed screws.4 Placement down the central axis permits insertion of a longer screw and distributes and reduces the bending forces more efficiently.5 Several investigators have attempted to quantify the optimum starting position for screw insertion to place the screw down the central one third of the proximal pole, which has been shown by Trumble and associates to be associated with faster union rates.6 Menapace et al defined a safe zone for volar K-wire insertion for placement of a Herbert-Whipple screw based on x-ray, CT, and anatomical dissections.7 To prevent potential injury to the scaphoid blood supply they recommended avoidance of the radiodorsal portion of the scaphoid (70 to 80% of the blood supply) and the volar surface of the scaphoid tuberosity (20 to 30% of the blood supply). They eschewed the ulnar third of the scaphoid so that the scaphocapitate articulation was not compromised. They also noted that there were no significant radiographic differences in the lengths and widths of paired scaphoids, which allowed them to use the contralateral scaphoid as a measuring template for deriving the optimal screw placement. Based on their findings they defined the safe starting point to be 9 to 23% of the radiographic length of the uninjured scaphoid moving in a dorsal and distal direction from the volar prominence of the tubercle as measured from the lateral x-ray, which in practical terms was 4 to 5 mm. The ideal targeting point for the screw point was 10% of the radiographic length of the contralateral scaphoid, which equated to 2 to 3 mm in a radial and slightly volar direction from the central part of the scapholunate interosseous ligament.

Leventhal et al performed a CT study of nine scaphoids that were randomly selected from a CT scan database at their institution. They computed a safe zone that was located 2.3 mm inside the original cortical bone surface, based on a 1.7 mm screw radius (Acutrak 2 Mini, Acumed LLC, Hillsboro, OR), a 0.035 mm cortical bone thickness, and an additional 0.25 mm safety margin. They found that the central axis of a 1.7 mm best-fit cylinder down the central axis was partially or completely obstructed by the trapezium in all of the specimens. When they looked at the longitudinal axis that allowed them to place the maximum-length crew, however, they found the safe zone passed on average 1.8 mm ± 0.8 mm away in primarily a dorsal and slightly radial direction (0.2 mm) from the apex of the scaphoid tubercle and exited ulnar and slightly dorsal to the longitudinal axis.8 They did acknowledge that placing the wrist in traction and ulnar deviation might provide greater access by further increasing the distance from the screw axis to the trapezium. Use of a larger-diameter screw would also have changed these findings.

Levitz and Ring attempted to quantify the guidelines for optimal scaphoid screw insertion for scaphoid waist fractures via a volar retrograde insertion technique with a Synthes cannulated-head screw (Synthes, Paoli, PA) by using quantitative computer analysis of CT images of 15 unfractured scaphoids.1 They defined the optimal screw placement as consisting of a 2 mm safe margin between the central axis of the screw and at least 5 mm of screw tip crossing the fracture site and engaging the apex of the proximal pole. They found that in all of the insertion planes there was a substantial concavity of the volar surface of the scaphoid that posed a problem for potential screw cutout. A more radial insertion point minimized this risk, but there was little margin for error because the average clearance of the trapezium measured 4 mm, which is the same diameter as most scaphoid screws. They also found that the most likely place for a volarly inserted screw to perforate the articular surface is the dorsal-radial surface of the scaphoid, which is exacerbated by any angular or humpback deformity ( Fig. 13.3 ). This region must be carefully evaluated fluoroscopically at the time of surgery by gradual pronation with the wrist in slight extension.

A relatively radial starting point facilitates placement of the screw tip in the center of the proximal pole and helps avoid the trapezium, but drilling or partial excision of the trapezium often may be necessary for optimal screw placement.