Fractures of the Carpal Bones

Biomechanics of Scaphoid Fractures and Implications of Nonunion

Although the exact mechanism of fracture is not completely understood, hyperextension past 95 degrees is the usual position of injury, but other mechanisms such as axial loading have also been postulated to produce scaphoid fractures, as has hyperflexion of the wrist. With the hyperextension mechanism, a fracture of the scaphoid usually begins at the volar waist with a tensile failure; the forces propagate to the dorsal surface with compression loading, until failure occurs. In a cadaveric study, wrists placed in extreme dorsiflexion and ulnar deviation produced fractures through the scaphoid waist as the scaphoid impinged on the dorsal rim of the radius. Proximal scaphoid fractures resulted from dorsal subluxation during forced hyperextension. Carpal dislocations and scapholunate ligament tears were reproduced with wrist extension and ulnar deviation, combined with intercarpal supination.

As with any fracture, the potential for healing relies on the fracture’s location, vascularity, and stability. Nonunion occurs in 10% to 15% of all scaphoid fractures. The risk of nonunion increases with:

• • 1.

    Delay of treatment for more than 4 weeks

  • 2.

    Proximal pole fractures

  • 3.

    Fracture displacement greater than 1 mm

  • 4.

    Osteonecrosis

  • 5.

    Tobacco use

  • 6.

    Associated carpal instability (dorsal intercalated segmental instability [DISI] with a scapholunate angle >60 degrees and a capitolunate angle >15 degrees) secondary to humpback (flexed with intrascaphoid angle >45 degrees; the normal intrascaphoid angle is 24 degrees) ( Fig. 16.5 )

    

For nondisplaced waist fractures treated with casting, nonunion rates are 5% to 12%. Nonunion rates for displaced scaphoid fractures treated nonoperatively are higher, reaching 50%.

Untreated displaced fractures of the waist are subject to varying degrees of these forces and will usually angulate as the volar bone is reabsorbed, yielding a “humpback” flexion deformity of the scaphoid. The resultant radial column shortening and the extension of the proximal scaphoid pole releases the lunate to rotate into DISI under the influence of the attached triquetrum. Ultimate treatment of a humpback scaphoid nonunion with DISI requires both restitution of scaphoid anatomy and reversal of the secondary changes in carpal kinematics.

Untreated scaphoid nonunion will predictably progress to arthritic change, in what has been termed scaphoid nonunion advanced collapse (SNAC). Arthritic change arises at the radial styloid articulation with the distal scaphoid pole (stage I) and is followed by degeneration of the scaphocapitate joint (stage II) and ultimately the midcarpal joint (stage III). Arthritic changes have been found in 97% of the patients assessed at least 5 years after injury, with the degree of arthritic changes being proportionate to the duration of nonunion. Patients generally present with escalating mechanical pain, with limitations in range of motion. Düppe and colleagues reviewed 30-year follow-up results of scaphoid fractures treated with thumb spica short-arm casts. Ten percent of the patients developed nonunion; 60% of these demonstrated radiographic evidence of radiocarpal osteoarthrosis, while only 2% of the healed group demonstrated degenerative change.

Clinical Presentation

The patient usually presents with pain on the radial side of the wrist. There may be swelling on the radial side as well. There is usually a history of trauma, such as falling on an outstretched hand, collision of the wrist against a person or heavy obstacle, or possibly a direct blow against an object. There may be limited range of motion and pain when applying extended wrist loading or positioning the wrist in extreme positions of flexion or extension.

Physical Examination

Physical examination starts with visual inspection. Wrists with acute fractures may have swelling and bruising in the radial aspect of the wrist. There may be limited range of motion. Wrists with chronic injury may have swelling in the dorsoradial wrist. “Snuffbox tenderness” has become synonymous with scaphoid fracture, but this applies predominantly to waist fractures, which represent 70% of scaphoid fractures. The second most common type of scaphoid fracture is a proximal pole fracture, at 20%. The least common is a distal pole fracture, at 10%. Fractures tend to occur at the waist partly because the RSC ligament acts as a fulcrum over which the scaphoid waist fractures ( Fig. 16.6 ).

The full physical examination of the scaphoid bone should include all of its parts: the waist, distal pole, and proximal pole. To palpate the anatomic snuffbox for the waist examination, palpate just distal to the radial styloid in the “soft spot.” The distal pole should be palpated at the scaphoid tubercle on the palmar aspect of the wrist. To do this, place the index finger in the anatomic snuffbox and place the thumb on the palmar aspect just distal to the anatomic snuffbox. The prominent bone palpated is the distal pole of the scaphoid. With radial deviation of the wrist, this prominence should move palmarly toward the examiner’s thumb. This distal pole should be checked for tenderness. The proximal pole is palpated dorsally in line with the second ray just distal to the dorsal radius lip. The scapholunate ligament is in line between the second and third rays just distal to the dorsal radius lip and corresponds to the 3-4 wrist arthroscopy portal. The proximal pole is just radial to the scapholunate ligament/3-4 portal area. Pain on longitudinal compression of the thumb (scaphoid axial compression test) is also a sign of scaphoid fracture. If all three tests of anatomic snuffbox tenderness, scaphoid tubercle tenderness, and scaphoid axial compression test are positive, there is 87% to 100% sensitivity and 74% specificity for scaphoid fracture. ^,

Diagnostic Imaging of Scaphoid Fractures

Radiographs are a mainstay in the workup of scaphoid fractures. Standard radiographic views are the posteroanterior (PA), lateral, oblique, and scaphoid views. It is important to obtain a true scaphoid pisiform capitate (SPC) lateral radiograph of the wrist, wherein the palmar cortex of the pisiform bone overlays the interval between the palmar cortex of the distal scaphoid pole and the palmar cortex of the capitate. This view allows a true assessment of carpal alignment.

A predictable scaphoid view is taken where a fist is made, with the thumb covering the dorsum of the middle phalanges of the index and middle fingers. The pronated forearm and hand are placed on the radiography table. The wrist is placed in ulnar deviation ( Fig. 16.7 ). The rationale behind this position is to take the scaphoid out of its usual position of flexion and pronation. With the thumb in the above position, the wrist is extended and supinated slightly. Ulnar deviation and wrist extension extend the proximal carpal row. This view allows a full view of the scaphoid bone with minimal overlap from neighboring bones.

A clenched pencil view can also be useful. ^, This is the best view to assess associated dynamic scapholunate widening ( Fig. 16.8 ) and also shows SNAC and scapholunate advanced collapse (SLAC) wrist changes better than standard PA views. For clinically suspected occult scaphoid fractures with negative initial radiographs, the wrist may be casted and radiographs repeated in 10 to 14 days.

Computed Tomography

Computed tomography (CT) scanning helps elucidate scaphoid fracture displacement, bony morphologic findings, gapping, sclerosis, cysts, and evidence of healing. CT is particularly helpful in addressing nonunions. It is important that CT scans are taken with overlapping 0.625 mm cuts along the long axis of the scaphoid and with coronal and sagittal reconstructions. CT scans have particular utility in evaluating for healing after scaphoid surgery, especially because radiographs are often indeterminate. CT scanning has also demonstrated some utility in evaluating osteonecrosis of the proximal pole of the scaphoid. Increased radiodensity of the proximal pole is a sign of dysvascularity.

Gilley and colleagues showed that radiographs frequently underestimate fracture displacement compared with CT scan. Thirty-nine preoperative radiographs and CT scans were evaluated by two blinded readers. Of these, 26% to 33% of nondisplaced fractures on radiographs were read as displaced on CT scan. If the status of the radiocarpal or midcarpal cartilage is in question, a magnetic resonance imaging (MRI) without contrast with cartilage-sensitive sequencing may be obtained.

Scaphoid healing cannot be reliably determined by standard radiographs at 3 months ; consequently, CT provides enhanced resolution and definitive information regarding healing.

Magnetic Resonance Imaging

MRI may be helpful to determine whether there is occult scaphoid fracture, especially in the acute period after injury. Specificity is 90% and sensitivity is between 90% and 100%, as opposed to bone scintigraphy (i.e., bone scan), which is 92% to 95% sensitive and 60% to 95% specific. Karl and colleagues performed a study comparing cast immobilization and repeat radiographs at 2 weeks, immediate CT scan and immediate MRI for occult scaphoid fractures. They concluded that advanced imaging for suspected scaphoid fractures in the setting of negative radiographs represents a cost-effective strategy for reducing both morbidity and costs. The decision of use of CT versus MRI is a function of local costs and test performance capabilities. In a comparison of CT and MRI scans for suspected scaphoid fracture, both modalities had similar characteristics. The sensitivity, specificity, and accuracy were 67%, 96%, and 91%, respectively, for CT and 67%, 89%, and 85%, respectively, for MRI which were not significantly different.

Bone Scintigraphy

Bone scintigraphy can be used in cases of clinically suspected scaphoid fractures. In a Cochrane study by Mallee and colleagues, bone scintigraphy was compared with MRI and CT scans in patients with negative radiographs. After 72 hours from injury, bone scintigraphy was statistically the best diagnostic test; however, it has the drawback of increased radiation exposure and may not be available at all institutions.

Stable Acute Fractures

Scaphoid Fractures in Children

Fractures of the immature scaphoid are uncommon and can be challenging to diagnose. These fractures most commonly involve the distal scaphoid and are effectively treated with cast immobilization. Fortunately, most acute pediatric scaphoid fractures heal with nonoperative treatment, and scaphoid fracture fixation in children is indicated only if there is nonunion.

The diagnosis of acute scaphoid fracture may be missed or delayed because of minimal symptoms. This is particularly true in athletic adolescents, who may return to sports prematurely after a seemingly minor injury to the wrist. The presentation of scaphoid fractures in adolescents has changed over the years and today more closely resembles the adult pattern. Malunion or nonunion may occur in patients with a missed diagnosis or delayed presentation and occasionally in patients treated promptly with immobilization. Surgical intervention should be considered for fracture nonunions in patients who are at or near skeletal maturity or in those in whom nonsurgical treatment has failed. Mintzer and Waters presented the outcome of 13 pediatric scaphoid fracture nonunions in 12 children treated over an 18-year period. The average elapsed time between fracture and surgery was 16.7 months. Four of the nonunions were treated with inlay bone graft from a palmar approach, and nine were treated with Herbert screw fixation and iliac crest bone grafting. The average time of follow-up was 6.9 years (range, 2 to 19 years). All cases went on to clinical and radiographic union. Patients had no statistically significant difference in range of motion or strength between the operative and contralateral normal wrists. The length of time for postoperative immobilization in the Herbert screw group was significantly less than that in the inlay bone graft group. Though scaphoid nonunions in children can heal with prolonged cast immobilization, the authors recommended that the treatment of scaphoid fracture nonunions in the skeletally immature patient be rigid fixation with a compression screw and iliac crest bone graft.

Masquijo and Willis presented their series of 23 pediatric (average age, 15 years) scaphoid nonunions fixed with iliac crest bone graft and screws, with a 95.6% union rate.

Weber reported on six children with nonunited scaphoid fractures treated nonoperatively. The mean age was 12.8 years (range, 9.7 to 16.3 years), and the mean follow-up time was 67 months. Five had no previous treatment, and the time to diagnosis averaged 4.6 months (range, 3 to 7 months) after injury. Treatment consisted of cast immobilization until clinical and radiologic union. Fractures united in all six children after a mean period of immobilization of 5 months (range, 3 to 7 months). All patients returned to regular activities. Although prolonged treatment with cast immobilization resulted in union of the fracture and an excellent subjective wrist score in all patients, this delay may not be well tolerated by the child or family.

Delayed (Subacute) Presentation of Waist Fracture

Surgery is generally indicated for delayed presentation of a scaphoid fracture 4 to 6 weeks or more following injury, even in seemingly nondisplaced fractures. Patients with delayed presentation of scaphoid fractures have a higher likelihood of nonunion with closed treatment and require 4 to 6 months to heal in plaster.

Managing Scaphoid Fractures in Athletes

Scaphoid fractures are common in competitive and recreational athletes, and such patients are reluctant to submit to the long period of immobilization and restricted activity that plaster requires. These factors may influence treatment decisions. The treating physician may permit participation in athletic activity with a playing cast. Some organizations permit casts that are protected with foam padding. Whether the forces generated by firm gripping are detrimental to healing probably depends on the fracture’s inherent stability. One study reported a faster return to play with internal fixation compared with a playing cast alone. A potential problem with plaster immobilization is compliance. Young people often will modify or remove the cast and are increasingly noncompliant with follow-up over time. The pressure to return to sports may lead the patient and coach to seek ways of shortening the period off of the field. The player’s ability to return to sports before the fracture is healed depends on the sport and its requirements. Options to be weighed will be surgery, a playing cast, and playing restrictions. The goal will be the successful union of the scaphoid fracture regardless of the patient’s athletic responsibilities. Clearly, educating the patient, family, trainer, and coach is essential. Every sport and different levels of the sport (professional, collegiate, high school) have different rules, and decisions are best customized based on individual circumstances. ^,

Mechanics of Fracture Fixation

Bone healing requires viable bone cells, an adequate blood supply, and stabilization of the fracture site. For a fixation device to be successful in providing rigid fixation of scaphoid fractures, it must be able to resist complex bending, shearing, and translational forces during normal functional loading. Because the majority of the scaphoid is covered with cartilage, fracture callus is not produced, so primary bone healing is entirely dependent on rigid stabilization of the fracture fragments until healing.

The mechanical effectiveness of internal fixation is determined by the bone quality, fracture geometry, fracture reduction, choice of implant, and implant placement. While all five of these independent variables are important, bone quality and fracture geometry are intrinsic to the patient. Fracture reduction, choice of implant, and implant placement are all under the surgeon’s control. Fracture reduction and placement of the implant in the biomechanically ideal position are the most important of the five variables.

Where and how to place the screw is controversial. Earlier literature focused on the longest central screw, where later literature focused on an awareness of the three-dimensional fracture geometry and placing the screw with all factors considered. While a central long screw is best for many fractures, it may not be best for all.

Trumble and colleagues observed that screws placed in the central third of the scaphoid were associated with significantly shorter healing times than screws placed outside of the central third axis ( P < .05). To explain this observation, McAdams and colleagues simulated scaphoid waist fractures and compared screws placed in the central axis with screws placed eccentrically. This study demonstrated that simulated fractures fixed with centrally placed screws in the proximal fragment demonstrated mechanical properties compared with screws placed in an eccentric position. Fixation with central screw placement demonstrated 43% greater stiffness, 113% greater load at 2 mm of displacement, and 39% greater load at failure.

Biomechanically, the longer the screw, the more rigid the fixation, because longer screws reduce forces at the fracture site and bending forces are spread along the threads. In a cadaveric study by Dodds and colleagues, short or long screws were placed along the central scaphoid axis after a simulated waist osteotomy, and constructs with longer screws were significantly stiffer than those repaired with short screws.

More recent literature focusing on the obliquity and location of the fracture line has called into question using a long central screw for all cases. Biomechanical modeling of preoperative CT scans demonstrated increased compression strength when cannulated screws were placed within 10 degrees of a perpendicular trajectory to the plane of the fracture. In another biomechanical study using preoperative CT scans, Kupperman et al. identified the ideal starting point and trajectory for proximal, waist, and distal pole fractures that predicted maximum “effective compression length,” a composite of screw perpendicularity and number of crossing screw threads.

When rigid fixation cannot be provided by a screw placement alone (as in extreme proximal pole fractures and nonunions), augmentation may be necessary to prevent micromotion at the fracture site. Supplemental fixation can be applied from the distal scaphoid to the capitate using a 0.045-inch or 0.062-inch Kirschner wire or a mini–headless screw. If a Kirschner wire is placed, it should be buried because it may be necessary to leave it in place for 3 months or until the bone is healed.

Implants for Rigid Fixation of Scaphoid Fractures

Implants used include Kirschner wires, headless compression screws, staples, and plates. Solid and cannulated screws are available from several manufacturers. Any implant used must reduce bending, shearing, and translational forces acting at a fracture site.

Kirschner Wires

Although Kirschner wires are easy to insert, they have a narrow role for scaphoid fixation today, given the relatively insecure fixation and minimal compression afforded by these implants. Kirschner wire fixation must be supplemented with a cast until healing, and a separate procedure for Kirschner wire removal is required. In multitrauma situations or open fractures, rapid stabilization of an unstable scaphoid fracture may be expedient.

Screws

In 1954, McLaughlin described fractures of the scaphoid as “an unsolved problem.” His main interest was returning a “breadwinner” to work with a treatment that would “hold bone fragments in apposition” until healing. He reported on the fixation of scaphoid fractures using solid lag screws. The operative procedure was technically challenging, the optimal screw position was not always achieved, and the incidence of nonunion in unstable fractures was not substantially reduced over that obtained by casting alone.

Herbert and Fisher in 1984 presented the results of the first headless screw used to treat 158 patients from 1977 to 1981. The rate of union was 100% for acute fractures and 83% overall. This screw revolutionized bone fixation because it permitted compression of a fracture with two heads of differential pitches ( Fig. 16.9 ). The embedded threaded heads of headless screws are placed in the densest bone of both poles for maximum bony purchase. This paper also demonstrated that screws placed perpendicular to an acute fracture plane using compression and rigid fixation could successfully heal an acute fracture. The implant, however, is not technically easy to insert, as it required a special compression jig that was difficult to position; other centers reported lower rates of union with screw fixation of acute scaphoid fractures secondary to technical problems.

The next major development in headless screw design was the cannulated compression screw. This device greatly simplified accurate placement of the screw within the scaphoid bone by using a thin guidewire placed under fluoroscopic control. A fully threaded, variable-pitch implant, the Acutrak screw (AcuMed, Hillsboro, OR), demonstrated compression comparable to that of a standard 4-mm compression screw and greater compression compared with the Herbert screw in biomechanical testing. Several manufacturers have developed cannulated compression screws with unique advantages and disadvantages. There have been a number of biomechanical studies comparing headless compression screws. Most second-generation headless compression screws achieve adequate compression; however, the compression decreases by 50% 12 hours after placement, and the compression does not necessarily correlate with the feel of torque on the screwdriver. What is unknown is how much compression is necessary for bone healing. In a biomechanical study of different screws in cadavera, no one screw demonstrated superior mechanical properties. Currently, screw selection is still largely a matter of surgeon preference.

Staple Fixation

The use of staples has had its proponents as a means of achieving stable fixation. Early studies demonstrated simple application and satisfactory healing rates among patients with scaphoid nonunions and acute fractures but long-term degenerative changes secondary to hardware impingement. New staples with memory achieve compression after insertion as they warm to body temperature. In a series of 131 patients with acute fractures and nonunions of the scaphoid waist, all acute fractures healed and all but 2 nonunions healed. However, pain was present in 21% of cases. This method has not replaced headless screw methods.

Dorsal/Mini-Open Approach ( Fig. 16.10 )

The main advantage of the dorsal approach is that it is technically easier to obtain an ideal screw position in the central axis of the scaphoid. Many surgeons prefer dorsal mini-open screw fixation because of the ease of access, the minimal soft tissue disruption, and the ability to place a screw down the central axis. ^, The disadvantage of the dorsal approach is that it requires flexion of the wrist, which may theoretically displace an unstable fracture. The dorsal mini-open approach for scaphoid fracture fixation is indicated for proximal pole and waist fractures. A small incision is safer than a purely percutaneous method when approaching from the dorsal wrist. Weinberg and colleagues have shown that there is a 13% chance of tendon injury with a purely percutaneous technique.

To perform the dorsal mini-open technique, a tourniquet is placed proximally on the arm and draped to allow full motion of the extremity. A 2- to 3-cm longitudinal or transverse incision is made over the proximal pole just distal to the Lister tubercle, and just radial to the area that corresponds with the 3-4 arthroscopic portal. The extensor pollicis longus (EPL) is carefully identified and the second and third dorsal compartment tendons are retracted radially. The capsule is incised off the radial rim and on the radial side of the Lister tubercle. If more exposure is necessary, the “window” approach is extended in an oblique fashion toward the lunate and just distal to the fibers of the dorsal radiocarpal ligament (see Chapter 13 and Video 13.4) .

Extreme care is taken to avoid disruption of the dorsal fibers of the SLIL and dorsal scaphotriquetral ligament when reflecting the capsular flap ( Chapter 13 and Video 13.4 ). Dissection in a plane tangential to the dorsal surfaces of the scaphoid and the lunate can be performed with a scalpel or Beaver blade. The distal boundaries of dissection of the scaphoid are determined by the vascular supply along the dorsal ridge, which must be preserved. Wide skin hooks are used to retract the capsule or the capsule is sutured to nearby soft tissue or skin. If the scaphoid is displaced, the hematoma should be evacuated and the scaphoid reduced with the aid of a dental pick or smooth Kirschner-wire joysticks, if necessary. The wrist is flexed and the entrance point on the scaphoid is identified 1 to 2 mm radial to the membranous portion of the scapholunate ligament and in the midportion of the scaphoid in the sagittal plane for long axis screw placement. For oblique fractures, preoperative CT scan is used to identify the optimal location of the starting point and screw trajectory to gain maximum effective compression length.

The surgeon should aim the guidewire for headless compression screw from the ideal starting point to the articular base of the thumb metacarpal (because the scaphoid distal pole, trapezium, and metacarpal base are collinear). Once this wire is placed, the surgeon must not extend the wrist because this will bend the wire at the radiocarpal joint and risk guidewire breakage. Fluoroscopic images are taken to assess the fracture and wire position. In order to obtain the PA images while keeping the wrist flexed, it is imperative to flex the elbow. An optimal position of the guidewire in all views (PA, lateral, and oblique) should be obtained. The optimal position is one that has a combination of being closer to perpendicular to the fracture yet has as much bone on either side of the guidewire as possible.

When the wire is ideally positioned, the wire is advanced to the subchondral bone on the distal end of the scaphoid bone and measured. The appropriate screw length is shorter than this distance by at least 4 mm. For an adult man, 20 mm is often an appropriate length. The screw should be relatively long but should definitely not be too long. If the screw is too long, it will either distract the fracture when it abuts the unyielding distal subchondral bone or protrude from the bone distally or proximally. Also, as the fracture is compressed the length of the bone will diminish. Once this length is determined, the surgeon may choose to drive the guidewire into the trapezium bone and out the thenar skin so that wire retrieval can be easily accomplished should the wire break during drilling or screw insertion. A second, antirotation wire may also be placed. The guidewire is overdrilled with a cannulated drill. Power is used for the near cortex, but then the rest is hand drilled because on power the drill can cut the guidewire, especially if there is a slight bend in the wire. If when drilling, the drill bit does not advance, then stop and get fluoroscopy to confirm that the wire is not bent and that the drill is collinear with the guide wire. Although some of the systems tout a “self-drilling” screw, drilling past the fracture can be performed as determined by fluoroscopy. After drilling, and depending on the particular screw system, countersinking the proximal hole can help avoid a catastrophic proximal pole fracture. Most systems come with a countersink. Particularly when using a tapered screw, countersinking will decrease the hoop stresses on the proximal pole to lessen the chance of fracture. After countersinking, a screw of the appropriate size is placed, making sure that it is well seated below the cartilage and into the proximal subchondral bone. A study by Hart and colleagues showed that the torque of the screw turns felt by the surgeon does not correlate with fracture compression. After the screw is placed, it is important to remove the guidewire and check all fluoroscopic views (PA, lateral, supinated, and pronated oblique views) to ensure an excellent fracture reduction and hardware position. Particularly helpful is the 60-degree pronated oblique view in demonstrated screw prominence from the proximal pole. If unsure about screw position, live fluoroscopy should be used, rotating the wrist 360 degrees. The capsule is closed with absorbable suture. The tourniquet is released, and hemostasis achieved. The skin is closed, and a well-padded short-arm splint is placed. Depending on fracture stability and patient compliance, a postoperative short-arm thermoplastic splint or short-arm cast may be placed at 2 weeks until the bone is healed; this may take up to 12 weeks following operation, especially for proximal pole fractures. Fracture healing is determined clinically by lack of tenderness and by imaging with radiographs. CT scan may be obtained at 3 months to confirm healing before releasing patients to sports and activities, provided their strength and motion is within 80% of their uninjured side.

Dorsal Percutaneous Technique

The main difference in the dorsal percutaneous technique is that it is a percutaneous technique; as such the entrance point is identified by flexing the wrist and identifying the superimposed rings on fluoroscopy of the proximal and distal poles. The guidewire is placed down the center of the superimposed rings. The wire may be advanced out of the radial aspect of the thumb so the wrist can be extended for standard wrist radiographs. The wire may be further withdrawn to perform percutaneous or arthroscopic reduction before reengaging the wire in the proximal fragment. When satisfactory scaphoid reduction and wire placement is confirmed, the wrist is again flexed and the wire driven dorsally for the remainder of the procedure. The skin about the wire insertion site is opened with a No. 11 blade and blunt dissection carried out down to the capsule prior to drilling and inserting the screw. Care must be taken to avoid injury to the extensor tendons in this technique.

Arthroscopy-Assisted Percutaneous Dorsal Scaphoid Fixation

The goals of arthroscopy-assisted stabilization of scaphoid fractures are to reduce displaced fractures without an open incision and provide secure fixation that will permit early motion until solid union has been achieved. The early results of arthroscopy-assisted percutaneous screw fixation of displaced fractures of the scaphoid suggest that minimally invasive reduction and fracture union can be predictably obtained with good to excellent functional results in the correctly selected patient. Avoidance of open exposure limits the potential for wrist ligament injury, may help preserve the blood supply, and minimizes postoperative stiffness. However, many scaphoid fractures are now treated with mini-open techniques that create little scarring and do not adversely affect postoperative motion.

Slade et al. reviewed his results in arthroscopy-assisted fixation from a dorsal approach in 27 consecutive patients. There were 18 waist fractures and 9 fractures of the proximal pole. Seventeen patients were treated within 1 month of injury, and 10 patients were treated late. All fractures healed, as documented by CT scan.

Arthroscopy-assisted fixation of scaphoid fractures also allows for simultaneous detection of associated intracarpal soft tissue injuries. Braithwaite and Jones originally reported on four patients with a fracture of the scaphoid with complete scapholunate dissociation. As in fractures of the distal radius, associated soft tissue lesions may occur with scaphoid fractures, and arthroscopic evaluation allows detection and management. It is not known whether early arthroscopic detection and management of the associated injuries improve the final outcome. For operative details of arthroscopic-assisted percutaneous treatment of scaphoid fractures and scaphoid nonunions, the reader is referred to Chapter 17 .

Open Palmar Approach

The palmar approach may be used for acute waist fractures, scaphoid nonunions of the waist with humpback deformity, and the rare distal pole fracture that may require surgical fixation. An advantage of the palmar approach is that the wrist is extended, theoretically helping to reduce the fracture. A disadvantage is that it is not possible to place the screw down the true axis of the bone because the trapezium blocks the entrance of the center of the distal scaphoid ( Fig. 16.11A ).

The placement of the screw is always oblique; the more proximal the fracture, the less chance of engaging sufficient screw threads in the proximal fragment. Another concern is penetration of the dorsal surface of the scaphoid bone by the screw, especially when attempting to capture more proximal fractures. The optimal radiographic views for determining dorsal screw prominence are the 60-degree pronated oblique view as reported by Kim and colleagues ( Fig. 16.12 ).

The open palmar approach is most commonly used for scaphoid nonunion and less frequently for displaced comminuted or flexed acute scaphoid fractures. In the open volar approach, a hockey-stick incision is made over the flexor carpi radialis (FCR) tendon in the distal forearm and angled across the distal wrist crease toward the base of the thumb. The FCR tendon is retracted ulnarly and the radial artery radially. The wrist is hyperextended over a bump. The wrist capsule is entered through a curvilinear incision from the volar lip of the radius to the proximal tubercle of the trapezium. The capsule and intracapsular ligaments are carefully divided and reflected sharply off the scaphoid with a scalpel. The capsule needs to be preserved, as it contains the RSC ligament and will be repaired at the close of the procedure. Some surgeons prefer to tag the ligaments at this stage with nonabsorbable sutures for later repair. The entire volar scaphoid is exposed. Reduction is performed with dental pick or by manipulation or with joysticks. Bone grafting can be performed as required for volar comminution or in subacute fractures, with the grafts harvested from the volar radius beneath the pronator quadratus by extending the incision proximally an additional 2 to 3 cm. The scaphotrapezial joint may be opened to place a guidewire more centrally in preparation for final fixation. If necessary, part of the proximal trapezium can be excised with a rongeur to clear an unobstructed path for the guidewire and screw. Rigid internal fixation can be performed with the screw of choice.

With the volar approach, it is important to know that it is impossible to put the screw down the long axis of the scaphoid. Verstreken and colleagues have popularized a transtrapezial approach that enables more central screw position. ^, Leventhal and associates performed a computational analysis to identify the ideal starting point (1.7 mm dorsal and 0.2 mm radial to the tip of the scaphoid tubercle) to enable the longest possible screw trajectory in the scaphoid without violating the trapezium.

Palmar Percutaneous Method

Percutaneous scaphoid fixation may be performed with or without traction through a palmar approach. Fractures suitable for treatment with the volar percutaneous (distal to proximal) approach are fractures at the waist and distal pole. Humpback deformities or scaphoid collapse with a DISI deformity usually require open reduction, bone grafting, and fixation. Proximal pole fractures are best treated via a dorsal (proximal to distal) approach. An advantage of the palmar approach for percutaneous fixation is that the entrance of the scaphoid tubercle is subcutaneous; thus there are no tendons in the path of the approach. The incision may be enlarged slightly as needed to allow overdrilling of the guidewire and placement of the screw. Like the dorsal approach, a percutaneous volar approach can be extended if necessary to openly reduce the fracture or treat a nonunion.

Streli was the first to describe volar percutaneous screw fixation of the scaphoid fracture in 1970 using traction applied through the thumb and a standard Association for the Study of Internal Fixation (ASIF) screw. In 1991 Wozasek and Moser reported on an adaptation of Streli’s technique using cannulated 2.9-mm screws via a volar percutaneous approach. Later, Bond and colleagues demonstrated faster healing and return to work with modern cannulated headless compression screw fixation from the palmar approach, compared with cast immobilization.

To perform percutaneous volar scaphoid screw fixation, the patient is placed supine on an operating table and the hand is suspended vertically by the thumb using a finger trap ( Fig. 16.13 ). This position extends the scaphoid and ulnarly deviates the wrist to improve access to the distal pole of the scaphoid. A fluoroscopic imaging unit is rotated parallel to the floor and positioned so that the wrist is in its central axis. Traction from a tower permits full rotation of the scaphoid in the imaging beam. In the majority of cases, longitudinal traction is enough to reduce the scaphoid fracture. If the fracture is not reduced, Kirschner wires can be inserted and used as joysticks to manipulate the fragments into position. The quality of the reduction can then be checked radiographically.

Having achieved an acceptable reduction, the most important step is to establish the entry point of the guidewire. The ulnar deviation of the wrist extends the scaphoid to make the tubercle more accessible. The entry point, the scaphoid tuberosity, is located using a 12- or 14-gauge intravenous needle introduced on the anteroradial aspect of the wrist, just radial and distal to the scaphoid tuberosity. The needle serves as a trocar to guide the wire and establish a central path along the scaphoid. The needle is inserted into the scaphotrapezial joint and tilted into a vertical position. The needle is levered on the trapezium, which brings the distal pole of the scaphoid more radially and facilitates screw insertion. The wrist is rotated in the fluoroscopic beam to confirm that the needle is aligned along the axis of the scaphoid in all planes, with the intent of directing the guidewire into the proximal pole to a point just radial to the scapholunate ligament. Leventhal and colleagues identified the ideal starting point to be approximately 2 mm dorsal and just radial to the apex of the scaphoid tubercle in order to achieve maximum guidewire length within the scaphoid.

Once the entry point and the direction of the guidewire are confirmed, the needle is impacted into the soft articular cartilage over the distal pole of the scaphoid so that the tip does not slip during insertion of the guidewire. The guidewire is passed down through the needle and drilled across the fracture; its direction is continually checked on the image intensifier and adjusted as necessary, with the goal of entering the radial aspect of the proximal pole. The position is checked in multiple fluoroscopy planes, and, if satisfactory, a longitudinal incision of 0.5 cm is made at the entry point of the wire and deepened down to the distal pole of the scaphoid using a small hemostat and blunt dissection. This is a relatively safe zone, with minimal risk to the adjacent neurovascular structures.

The length of the screw is determined using a depth gauge or by advancing a second guidewire of the same length up the distal cortex of the scaphoid and subtracting the difference between the two. The correct screw size is 4 to 5 mm shorter than the measured length, which will ensure that the screw head is fully buried below the cartilage and the subchondral bone on each end. In rare cases a second antirotation wire may be inserted parallel to the first prior to drilling and reaming. The wire is advanced into the radius or out of the dorsal skin and clamped so as to avoid loss of wire position with drilling. The 12-gauge needle is removed and the cannulated drill is passed over the wire and advanced under imaging guidance, stopping 1 to 2 mm short of the far articular surface. At this point the hand is taken out of traction so that the screw will adequately compress the scaphoid. A countersink or a trailing drill is used, depending on the particular set. A self-tapping screw is advanced over the guidewire. The final position is checked with multiple fluoroscopic views to confirm complete containment within the scaphoid. A hyperpronated PA view profiles the dorsal-radial cortical margin of the scaphoid, where a perforation of the proximal cortical bone can occur. Compression of the fracture site is confirmed radiographically on the image intensifier. The wire is removed, the skin closed with a suture or Steri-Strips, and the wound covered with a sterile compressive dressing.

Combined Fractures of the Scaphoid and Distal Radius

Combined fractures of the distal radius and scaphoid are uncommon but present a challenging treatment dilemma. Scaphoid fractures may not be recognized when associated with a comminuted distal radius fracture and when untreated can result in carpal collapse, cystic degeneration, and eventual carpal degenerative arthritis. Although an isolated stable scaphoid fracture might be safely managed with plaster immobilization, the 12 to 16 weeks of immobilization required for healing are not appropriate for the treatment of the distal radius fracture. Prolonged immobilization may result in arthrofibrosis and atrophy of the forearm and hand, making the recovery of full hand function challenging.

A review of the published reports on combined scaphoid and distal radius fractures demonstrates that treatments have evolved over the past decade. Trumble and colleagues reported on six patients treated with internal fixation for ipsilateral combined fractures of the scaphoid and radius. All patients sustained a high-energy injury from a fall from a substantial height. All of the fractures united, with the radial fractures healing in an average of 6 weeks and the scaphoid fractures healing in an average of 13 weeks. Internal fixation of the scaphoid in these combined injuries allowed for earlier and more aggressive therapy to maximize wrist and forearm motion. My preferred treatment for combined distal radius fractures with scaphoid fracture is ORIF of the distal radius fracture and the mini-open approach and fixation of the scaphoid fracture with a headless compression screw. (For additional information, see Chapter 15 .)

Transscaphoid Perilunate Fracture-Dislocations

Acute fracture-dislocations of the carpus are uncommon. Perilunate fracture-dislocations represent approximately 5% of wrist fractures and are about twice as common as pure ligamentous dislocations. Transscaphoid perilunate fracture-dislocation is the most common type of complex carpal dislocation.

Treatment of these injuries can be challenging owing to the extensive soft tissue, cartilaginous, and bony damage. Furthermore, obtaining universally excellent long-term results can be elusive. Various operative treatment options have been recommended, including dorsal, volar, percutaneous, and arthroscopic approaches.

These injuries are usually due to a high-energy impact, as may occur in motor vehicle accidents, a fall from a height, or contact sports. The mechanism of injury characteristically involves forceful wrist extension, ulnar deviation, and intercarpal supination. The injuries have been classified as greater and lesser arc injuries and can take many forms (for more information, see Chapter 13 ). My preference is to get CT scans for all perilunate injuries, as some greater arc injuries can be subtle and overlooked on plain radiographs.

Herzberg and Forissier investigated the medium-term results (mean follow-up, 8 years) of a series of 14 transscaphoid dorsal perilunate fracture-dislocations treated operatively at an average of 6 days following injury. Eleven underwent ORIF through a dorsal approach. Combined palmar and dorsal approaches were used in three cases: in two cases, ORIF, and in one case, proximal row carpectomy. All internally fixed scaphoids healed, and no carpal AVN or collapse was observed. Carpal alignment was satisfactory in most cases. Posttraumatic radiologic midcarpal arthritis or radiocarpal arthritis, or both, was almost always observed.

Nearly every combination of radiocarpal and intercarpal dislocation has been described, but few fit neatly into a particular pattern or classification scheme. Although arthroscopic techniques and fluoroscopically aided percutaneous techniques have been described, my preference for treatment is an open procedure. It is imperative to specifically ask the patient about median nerve symptoms and perform physical examination for it prior to surgery. If left untreated, median nerve compression can lead to long-term nerve deficits, hand stiffness, and, potentially, complex regional pain syndrome.

The wrist is approached dorsally primarily for treatment of the scaphoid fracture or the scapholunate tear if it is a lesser arc injury. In 3% of cases, a complete SLIL tear accompanies a scaphoid fracture-dislocation. An extended carpal tunnel approach is added if one of two situations is present: (1) The preoperative history and examination are consistent with acute carpal tunnel syndrome, or (2) a perilunate fracture-dislocation is irreducible from the dorsal approach. The lunate is visualized in the space of Poirier between the RSC ligament and long radiolunate (LRL) ligament in an extended carpal tunnel volar approach to the carpus. Using a Freer elevator or similar instrument from the volar approach and manual wrist manipulation, the lunate can be negotiated into reduction.

For associated lunotriquetral ligament injury, the lunotriquetral joint is reduced and held with two Kirschner wires. One technique that is very valuable and surgically efficient is to place one or two double-ended Kirschner wires from within the lunotriquetral joint out the ulnar side of the wrist through the triquetrum before the lunate is reduced. A similar technique can be used in the scaphoid when the SLIL is torn. It is important to protect the soft tissues (radial artery, superficial radial nerve, tendons, veins) on the radial side of the wrist. Smooth Kirschner wire drilling in oscillation mode can be helpful to decrease risk of injury to soft tissues. The lunate is reduced and the wires are driven back into the lunate from the triquetrum and scaphoid, respectively. The Kirschner wires are buried since they must remain for 12 weeks. Nonburied Kirschner wires have a higher incidence of infection if they are left in more than 6 weeks. If extended carpal tunnel release has been performed, the stronger palmar component of the lunotriquetral ligament can be sutured with nonabsorbable braided suture from the palmar side.

Occasionally the dorsal radioulnar ligament origin is avulsed off of the dorsoulnar corner of the radius. The distal radioulnar joint will be unstable and the dorsoulnar corner of the radius will be devoid of soft tissue during the exposure. It is important to repair the ligament with bone suture anchors. Repairs are generally protected with cast immobilization for 8 to 12 weeks, followed by removal of Kirschner wires and then gentle active mobilization of the wrist over the next 4 to 8 weeks ( Fig. 16.17 ).

Evaluation of Scaphoid Nonunion

When evaluating patients with scaphoid nonunion, the following issues regarding the fracture must be considered:

• • 1.

    Where is the nonunion? At the waist or at the proximal pole?

  • 2.

    Is the nonunion displaced?

  • 3.

    Is there a humpback deformity?

  • 4.

    Is there comminution, cyst formation, or cavitation?

  • 5.

    Is DISI present?

  • 6.

    Was there previous surgery?

  • 7.

    Is the proximal pole dysvascular?

  • 8.

    Is the proximal pole fragmented?

  • 9.

    Is there arthritis (SNAC wrist)? If so, at what stage?

  • 10.

    Is the scaphoid deformed or salvageable?

Classification

Several classification schemes have been proposed for scaphoid nonunion, generally based on factors such as time since injury, mobility of the fragments, cystic or flexion deformity, and degenerative change. Two such classification systems are listed in Tables 16.2 and 16.3 . Table 16.4 presents a four-part practical classification and treatment algorithm based on fracture nonunion characteristics and vascularity.

Type IIB: scaphoid waist nonunion, deformity without arthritis

The most common type of nonunion faced by the surgeon will be a midwaist fracture nonunion with fracture site resorption and varying degrees of deformity and bone loss. For this type, nonvascularized autogenous bone graft will generally suffice for healing of the bone as long as certain technical aspects are followed. A point of controversy is whether deformity must be corrected to provide an optimal outcome. In a long-term follow-up study of healed scaphoid nonunions, Jiranek and colleagues identified a higher percentage of arthritic changes in malunited scaphoid fractures. The degree of arthritis did not correlate with the activity level, satisfaction, or return to work, however. These findings were confirmed in a recent study demonstrating periscaphoid arthritis in 10 of 22 patients with asymptomatic scaphoid malunion. Although anatomic alignment is intuitively preferable for restoration of normal kinematics, some degree of malunion seems to be tolerable. ^{, ,} It is therefore recommended to correct intercarpal malalignment and DISI at the time of nonunion surgery, especially in young and active patients, to avoid the otherwise probable progression of arthritis.

Surgical fixation with nonvascularized local autogenous bone graft

The procedure can be done from a volar or dorsal approach, depending on the location of nonunion. Generally, midwaist nonunion surgery is performed from the volar approach and proximal pole nonunion surgery is performed from the dorsal approach. Careful preparation of the bone fragments so as not to cause thermal necrosis with power tools is imperative. Removal of sclerotic and dysvascular bone with curettes to expose healthy bone surfaces is advantageous for both proximal and distal poles. The scaphoid is anatomically reduced and abundant autogenous bone graft packed into the proximal and distal poles. Fixation with a headless compression screw is optimal and is generally performed from distal to proximal in waist nonunions, as detailed in the Hybrid Russe Procedure section. When the lunate is malrotated into dorsiflexion (DISI), it often corrects when the scaphoid’s humpback deformity is reduced. However, if DISI is not corrected with scaphoid reduction, then reduction and temporary fixation of the lunate with a temporary 0.062-inch radiolunate Kirschner wire will facilitate accurate alignment of the scaphoid proximal pole prior to scaphoid grafting and fixation (see Fig. 16.16 ). Supplemental fixation across the scaphocapitate joint with a Kirschner wire may further stabilize the construct, especially if there is concern about the rigidity of fixation.

For nonvascularized bone graft, healing rates are identical between cancellous only and corticocancellous grafts, but cancellous only grafts demonstrated a statistically quicker time to union. There is no difference in healing or time to union between graft sources (iliac crest versus distal radius), ^, but the iliac crest graft has the disadvantages of another operative site, longer surgery time, more pain, and risk for cutaneous nerve injury.

In terms of fixation, a systematic review of the literature showed the superiority of screw over Kirschner wires for fixation of the scaphoid. Cohen and colleagues reported on 12 patients who all healed with cancellous bone graft from the distal radius and a distal to proximal headless compression screw. The average lateral intrascaphoid angles were improved from 49 degrees to 32 degrees.

Several authors have reported on the use of plate fixation for scaphoid nonunions. Huene and Huene reported their experience with plate fixation to stabilize scaphoid fractures in 1991. The Ender blade plate is suitable for adding stability in scaphoid nonunions with AVN, cystic degeneration, and osseous size discrepancy or compromise. Plate fixation can be used with scaphoid nonunion humpback deformity. Dodds and colleagues presented a combination of volar buttress plating and a vascularized volar distal radius wedge autograft pedicled on the volar carpal artery for salvage treatment of chronic scaphoid nonunion. Ghoneim reported on healing of 13 of 14 scaphoid nonunions with an iliac crest wedge graft and a volar miniplate of 1.5 or 2 mm. Other authors have had had unsatisfactory results, however. In a series of 15 patients with scaphoid nonunion, there was 87% union but a high percentage of hardware complications and need for secondary surgery for hardware removal in one-third of patients.

Matti technique, Matti-Russe technique, and Green’s modification of the Russe technique

In the original Matti technique, only cancellous bone was used. The Matti-Russe technique was a cancellous plug from the iliac crest and cancellous chips, and Green’s modification of the Russe technique was two back-to-back corticocancellous intramedullary iliac crest strut grafts and cancellous chips ( Fig. 16.18 ). Green reported 24 of 25 (92%) with good proximal pole vascularity united. If proximal pole vascularity was diminished, then union was 71% (10 of 14). If the proximal pole showed no bony bleeding, 0 of 5 united. In a long-term evaluation of Russe bone grafting, Barton had less success, especially in proximal pole nonunions, which demonstrated a union rate of only 54%.

Corticocancellous wedge graft for humpback deformity

In 1984 Fernandez presented his modification of the Fisk procedure to treat scaphoid nonunions associated with carpal instability. His article described the following modifications: (1) preoperative calculation of the exact scaphoid length and form based on comparative radiographs of the opposite wrist, (2) the use of a palmar approach, (3) the insertion of a wedge-shaped corticocancellous graft from the iliac crest after resection of the pseudarthrosis, and (4) the use of internal fixation ( Fig. 16.19 ).

Preoperative planning is considered essential to restore the anatomic length, analyze the angular deformity, evaluate the pathologic scapholunate angle, and calculate the resection and size of the graft needed. The palmar approach reduces the danger of iatrogenic damage of the vascular supply of the scaphoid and accidental lesions of the superficial branches of the radial nerve. Iliac bone is preferred to the radial styloid graft, as proposed by Fisk, because of its better ability to resist compression forces. Internal fixation adds rotational stability so that continued postoperative plaster immobilization can be reduced to a minimum of 8 weeks.

Technique

Preoperative PA, lateral, and ulnar deviation PA radiographs of the injured wrist and the opposite normal wrist are used to determine the amount of resection and size of the graft needed. The surgeon must determine the amount of scaphoid flexion deformity, the extent of carpal collapse, and the scapholunate and lunocapitate angles ( Fig. 16.20 ). The humpback deformity can be evaluated more precisely using CT scans reformatted along the long axis of the scaphoid.

The nonunion is approached through a palmar hockey-stick incision along the FCR, extended obliquely toward the thenar eminence. The anterior capsule is incised from the distal radius to the scaphotrapezial joint to expose the volar surface of the scaphoid. The capsular flaps contain the RSC and LRL ligaments. Sclerotic or irregular borders of the nonunion site are resected with an oscillating saw to obtain flat bony surfaces. Cystic defects are curetted and filled with cancellous bone. The extended lunate is corrected by flexing the wrist until the lunate assumes a neutral position under fluoroscopy, and a 0.062-inch Kirschner wire is placed through the diaphyseal-metaphyseal junction of the radial styloid into the lunate under fluoroscopic control ( Fig. 16.21 ). The flexion deformity and shortening of the scaphoid are then corrected by distracting the osteotomy with small bone hooks or large skin hooks while hyperextending the wrist over a rolled towel. A bicortical or tricortical graft is harvested from the iliac crest, along with additional cancellous bone. The graft is kept in aspirated blood from the iliac wound until ready for use. Osteotomes of various sizes can be used to measure the depth, width, and length of the trapezoidal defect of the scaphoid, and the graft is sculpted to the correct dimensions using a saw and osteotomes. The bone graft is impacted into the defect, and scaphoid alignment and posture are checked with fluoroscopy. A guidewire is advanced from distal to proximal across the scaphoid and intercalated graft. Finally, compression screw fixation of the scaphoid is performed. Careful closure of the palmar capsule and repair of the RSC and LRL ligaments are done. A long-arm thumb spica postoperative splint is applied. Sutures are removed at 10 to 12 days, and the above-elbow immobilization is continued until the radiolunate pin is removed.

For long-standing DISI deformity that presents with a radiolunate angle greater than 15 degrees, the radiolunate transfixion wire is left in place for 4 to 6 weeks to reduce forces on the healing scaphoid. Short-arm casting is continued after wire removal until radiographic or CT evidence of trabecular healing has occurred. Fernandez’s criteria for establishing healing include the absence of pain, radiographic evidence of bridging bony trabeculae on both sides of the interposed graft, a disappearance of the osteotomy lines in conventional x-rays, and no signs of screw loosening. CT is advised to confirm union prior to return to athletic activities.

Outcomes

In 1990 Fernandez reported union of 19 of 20 established nonunions repaired with this technique, with an average time off work of 8.9 weeks. In a review of the pertinent literature, he determined that the most common reasons for recalcitrant nonunion were improper internal fixation techniques or the absence of bone grafting, or both.

Eggli et al. reported on a retrospective review of 37 patients with scaphoid fracture nonunions treated by interpositional bone grafting and internal fixation at an average follow-up of 5.7 years. Solid radiographic union was achieved in 35 cases. Preexisting AVN was an adverse factor for achievement of union and satisfactory outcome. Patients with preexisting degenerative changes had a significantly worse clinical outcome. The vast majority of the patients had satisfactory correction of scaphoid length and the associated DISI. Although 30 patients showed radiographic evidence of mild or moderate degenerative changes at their latest follow-up, there was no significant progression of arthrosis, and carpal collapse deformity did not progress after healing of the fracture nonunion.

Arguments against the wedge graft

The wedge graft technique as described may not routinely incorporate curettage of the proximal and distal poles to healthy cancellous bone. Removal of sclerotic, fibrotic, or cystic bone in the proximal and distal poles prior to insertion of the wedge graft and replacing with additional cancellous graft may expedite union. Volar wedge bone grafting may require 3 to 6 months to heal and may result in reduced wrist motion. In cases without punctate bleeding, volar wedge grafting with an autologous iliac crest graft, Robbins and Carter demonstrated union rates of only 30%. This makes intuitive sense, as creeping substitution would be required to entirely revascularize the interposed corticocancellous graft prior to healing and revascularization of the proximal pole. Finally, the technique is exacting, and it may be challenging to simultaneously attain excellent compression, anatomic restitution of normal scaphoid anatomy, and rigid fixation.