Dorsal Approach

The patient is positioned supine on the operating room table with the arm abducted on a hand table. Guide wire insertion is accomplished free handed and percutaneously using fluoroscopic control, initially without tourniquet. I prefer antegrade screw insertion for proximal pole and waist fractures. The fluoroscopy unit is positioned over the arm board, parallel to the floor. The ARC traction tower (ARC Medical, Beaverton, OR) is ideally suited for this procedure because it has no central pole to obstruct instrumentation and the wrist can be flexed to 45 degrees in traction. This allows one to alternate fluoroscopic and arthroscopic assessment without moving the C-arm or extending the wrist and hence avoid the “dance of the K-wires” ( Fig. 17.1 ). Alternatively the wrist is flexed 45 degrees over folded towels, which places the scaphoid axis at 90 degrees to the beam and facilitates placing the screw down the central axis ( Fig. 17.2A,B ). I hand insert two K-wires into the mid-line of the scaphotrapezialtrapezoidal (STT) joint to act as targeting aids for guide wire insertion. The first guide wire is placed through the ulnar (STT-U) arthroscopy portal, which is located in line with the midshaft axis of the index metacarpal, just ulnar to the extensor pollicis longus (EPL) tendon. A second guide wire is inserted though a radial portal (STT-R), which is radial to the abductor pollicis longus (APL) tendon at the level of the STT joint. The targeting wires should intersect at the midpoint of the STT joint. The ideal starting point for the guide wire is at the most proximal tip of the scaphoid pole immediately adjacent to the insertion of the scapholunate interosseous ligament (SLIL). I place a third K-wire into the dorsal aspect of the scapholunate interval just ulnar to this point. Alternatively, with the scope in the 3,4 or 4,5 portal the tip of the guide wire can be directed to the soft spot on the proximal pole bordering the SLIL insertion.

Using a power drill, the guide wire is driven from an ulnar dorsal to a radial volar direction while keeping the wrist flexed. Aiming toward the intersection point of the dorsal STT K-wire on an AP fluoroscopic view guides the medial/lateral alignment of the guide wire. Pointing toward the intersection point of the radial STT K-wire on the semi-pronated fluoroscopic view guides the dorsal/volar alignment. The guide wire is then advanced distally through the trapezium and out through the skin ( Fig. 17.3A-E ). The wrist can be extended if necessary once the trailing end clears the radiocarpal joint. If the fracture is displaced, the guide wire is withdrawn distally until it lies solely within the distal fragment. Percutaneous 0.62 mm K-wires can then be inserted into the proximal and distal fragments and used as joysticks to align the scaphoid as described by Slade ( Fig. 17.4 ). The alignment of the concave scaphoid surface on the AP x-ray view can be used as a reference for fracture reduction. Once it is satisfactory, the reduction is captured by driving the guide wire proximally. A second antirotation K-wire should be inserted prior to reaming. The targeting K-wires are removed and the reamer is then introduced over the guide wire. Reaming stops 2 mm short of the distal pole ( Fig. 17.5 ).

The arm is exsanguinated and the tourniquet is elevated at this point. When using some other type of traction tower, the arm is suspended with 10 to 15 lb of traction with the wrist in extension. The quality of the fracture site reduction is visualized by inserting the arthroscope into the midcarpal ulnar (MCU) portal with the probe in the midcarpal radial (MCR) portal ( Fig. 17.5 ). Adjustments to the fracture reduction can be performed by withdrawing the guide wire into the distal fragment and using a Freer elevator alternately inserted in the MCR and STT-U portals. Once the reduction is acceptable the guide wire is positioned within 2 mm of the STT joint with the wrist maintained in flexion. Most of the screw measuring guides from different types of headless screws overestimate the screw length; hence a second wire of equal length is placed percutaneously at the proximal scaphoid pole and parallel to the guide wire. The difference in length between the trailing end of each wire is the scaphoid length. The screw length selected should be 4 mm less than the scaphoid length. This permits 2 mm of clearance of the screw at each end of the scaphoid, thus ensuring complete implantation without screw exposure. If it is necessary to take the wrist out of traction for this part, the screw length can be gauged by driving the guide wire volarly and distally until the trailing end is in the subchondral bone of the distal scaphoid pole and the process is repeated. A recent anthropometric study by Bindra determined that the average scaphoid length in adult males was 31.3 mm +/− 2.1 while the average female scaphoid was 27.3 mm +/− 1.7; hence the longest screw lengths may range from 23 mm to 27 mm (see chapter 2).6

The guide wire is driven volarly once more so that it is left protruding both proximally and distally. This prevents guide wire migration during reaming and screw insertion. The wrist must remain flexed during this part, otherwise the guide wire will bend and block both reaming and screw insertion. The scaphoid is then power reamed to within 2 mm of the distal pole. Care is taken not to ream through the subchondral bone because this reduces compression along the fracture site. A headless screw is advanced under fluoroscopic guidance to within 1 to 2 mm of the opposite cortex ( Fig. 17.6 A-G ). Radiocarpal and midcarpal arthroscopy is now performed to check for screw cut-out. The rigidity of fracture fixation is assessed by palpating the fragments with a 1 mm hook probe or Freer elevator. After screw insertion, the guide wire is again driven distally if necessary to allow wrist extension, and the fracture site is inspected arthroscopically. If rigid fixation has not been achieved, Slade has recommended pinning the distal fragment to the capitate ( Fig. 17.7A-G ).7 This locks the midcarpal row and reduces fracture site motion, especially with short proximal or distal fragments where only a few screw threads cross the fracture site. If desired, percutaneous iliac bone graft harvested with a bone biopsy needle or demineralized bone matrix can be injected percutaneously by advancing the arthroscopic cannula down the guide wire and into the drill hole in the proximal scaphoid, prior to screw insertion ( Fig. 17.8A,B ).