Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Fractures and Dislocations of the Carpus
Fractures of the Scaphoid
Introduction
Almost exclusively, scaphoid fractures occur in young, active males. They account for 60% to 80% of all carpal bone fractures and are second only to fractures of the distal radius in the frequency of wrist fractures. The frequency of scaphoid fractures is approximately 21,500 per year in the United States. Nevertheless, the potential impact of these injuries is significant. The diagnosis may be challenging. The course of treatment is often protracted and even under the best of circumstances has a significant impact on these patients, many of whom are engaged in the most productive years of their lives. One European study demonstrated that the average time off work after a scaphoid fracture was 6 months. Even with proper treatment, at least 5% of these fractures fail to unite. There is a consensus that displaced fractures and fractures of the proximal pole of the scaphoid should be managed surgically. Controversy exists regarding the optimal treatment of nondisplaced scaphoid fractures.
Mechanism of Injury and Biomechanics
The scaphoid functions as a link between the proximal and the distal carpal rows via strong ligamentous connections, leaving the scaphoid waist susceptible to fracture. Two different mechanisms can produce fracture of the scaphoid. By far, the more common mechanism appears to involve hyperextension and bending. In elucidating the pathomechanics, Weber and Chao consistently produced scaphoid waist fractures in cadaver wrists by applying an axial load to the radial half of the palm while the wrist was stabilized in 95 to 100 degrees of extension. It is surmised that the proximal pole of the scaphoid becomes fixed in this position by the radius proximally, dorsally, and radially; by the capitate and lunate ulnarly; and by the long radiolunate ligament and the radioscaphocapitate ligament volarly. At the same time, the distal pole is free to translate dorsally with the distal row of carpal bones, which results in fracture, usually through the waist of the scaphoid. The palmar aspect of the bone fails in tension, and the dorsal aspect fails in compression. Smith and colleagues showed that osteotomy of the waist of the scaphoid in cadaver specimens causes a 27-degree volar angulation of the scaphoid, with consequent collapse deformity of the wrist with extension of the proximal row of carpal bones. This study confirmed the stabilizing link function of the scaphoid—between the proximal and the distal rows—that has been assumed by most investigators based on the anatomy of the scaphoid and on clinical experience. A recent computed tomography (CT) study of acute, displaced scaphoid waist fractures demonstrated proximal pole extension, supination, and volar translation, while the distal segment remained relatively stable. Although still indicative of the scaphoid's link between the carpal rows, this could have implications for reduction maneuvers in displaced fractures.
A less common mechanism of scaphoid waist fracture, termed the puncher's scaphoid, was documented by Horii and colleagues. In this instance, with the wrist positioned in neutral or slight flexion, force is transmitted axially along the second metacarpal, through the trapezium and trapezoid, resulting in a flexion moment through the distal pole of the scaphoid. The investigators noted a high incidence of open metacarpal head fractures in these patients. Wrist pain in this setting should therefore raise a high index of suspicion for scaphoid fracture.
Evaluation
Examination
A high index of suspicion is the key to early diagnosis. Scaphoid fractures may produce surprisingly little pain, swelling, or limitation of motion. Therefore if a young adult reports a history of a fall onto an outstretched hand and pain with tenderness in the anatomic snuffbox, the diagnosis of scaphoid fracture should be assumed until proven otherwise. Tenderness may be localized to the dorsum of the wrist or volar scaphoid tuberosity with proximal or distal pole fractures, respectively. Male gender, sports injury, anatomic snuffbox pain on ulnar deviation of the wrist within 72 hours of injury, and scaphoid tubercle tenderness at 2 weeks are independent predictors of fracture in patients with suspected fracture. With all four factors present, the likelihood of scaphoid fracture is 91%. Conversely, in this study, the lack of snuffbox pain on ulnar deviation of the wrist was 100% predictive of no scaphoid fracture. Other studies have also attempted to determine which physical examination findings on initial presentation are helpful to diagnose occult scaphoid fractures. Anatomic snuffbox tenderness appears to be the most sensitive, although a combination of examination tests including snuffbox tenderness, longitudinal thumb compression, and tenderness over the scaphoid tubercle may improve the specificity when initial radiographs are normal.
Imaging
Radiographic examination is still the best method for determining the presence of a fracture. Although many different views have been recommended, we find that a standard posterior-anterior (PA), an ulnar deviation PA, a true lateral (i.e., radius, lunate, capitate colinear), and a 45-degree pronation PA view make up a useful initial series ( Fig. 41.1 ). The lateral view is helpful in identifying carpal malalignment often seen in displaced scaphoid fractures. The ulnar deviation and the 45-degree pronation views position the long axis of the scaphoid more parallel to the radiographic plate and thus the fracture line more parallel to the radiograph beam. Also, ulnar deviation tends to distract the fracture fragments (see Fig. 41.1D ). Both of these factors facilitate visualization of the fracture line. If these radiographs are equivocal or negative in the face of strong clinical suspicion, further oblique views can be taken. Stecher recommended a PA view that is angled 20 degrees from the vertical from distal to proximal. This too has the effect of directing the radiograph beam more parallel to the fracture line.
The false-negative rate for initial radiographs is reportedly between 2% and 25%. If clinical suspicion for a scaphoid fracture remains high, one can either begin empiric cast treatment or perform additional imaging studies, such as a CT, magnetic resonance imaging (MRI), or bone scan. If empiric treatment is chosen, the patient should be placed in a thumb spica cast for 2 weeks and the examination repeated. Although an occult fracture would be expected to become apparent on subsequent radiographs as resorption at the fracture site occurs, follow-up plain radiographs are, in general, of little value. Controversy exists as to whether MRI or CT is the better imaging study to evaluate occult scaphoid fractures. In one Swedish study of 300 wrists with radial-sided wrist pain and normal radiographs, MRI demonstrated a total of 224 fractures. Of these, 42% were interpreted as having scaphoid fractures, and 15% were distal radius fractures. However, false positives do occur. In addition, it is unclear what the significance of a “bone bruise” is. A recent small study comparing CT and MRI concluded that they are equivalent for evaluating suspected scaphoid fractures, with sensitivity, specificity, and accuracy of 67%, 96%, and 91%, respectively, for CT, and 67%, 89%, and 81%, respectively, for MRI. The authors concluded that both studies are better at excluding scaphoid fractures than they are at confirming them. Unresolved issues include the significance of “bone bruising” on MRI and unicortical fracture lines on CT.
A distinct advantage of CT is the ability to better define the fracture and assess for displacement. When ordering the CT, it is essential to ask the radiologist for images in the sagittal plane of the scaphoid ( Fig. 41.2 ) because otherwise axial views of the wrist are routinely done, and these are difficult to interpret. In other words, the same principles apply to CT views as to plain views. To be most useful, the beam must be parallel to the suspected fracture line or perpendicular to the long axis of the scaphoid. If axial cuts have been obtained, software is available to convert the images to be collinear with the scaphoid (OsiriX Imaging Software, www.osirix-viewer.com ). Images should be obtained at 1-mm intervals. Additional information about intercarpal alignment can be gained on the lateral (sagittal) views because the radius–lunate–capitate relationship is well visualized, as is fracture angulation within the scaphoid (see Fig. 41.2C ). In our practice, MRI is reserved for those patients in whom a significant intraarticular injury, aside from a scaphoid fracture, is suspected.
In managing suspected occult scaphoid fractures, immediate MRI or CT may be more cost-effective than casting for 2 weeks followed by repeat radiographs. This may be particularly important for patients who would be unable to work in a cast (i.e., athletes, laborers). However, no study to date has proven that immediate diagnosis of occult scaphoid fractures with advanced imaging improves outcomes compared with empiric casting and repeat imaging 2 weeks later.
Diagnosis and Classification
Although no single, universally accepted classification of scaphoid fractures exists, guidelines ( Figs. 41.3 and 41.4 ) exist to help the surgeon design the best course of treatment for a given patient. The fracture can be defined according to several features.
Duration
Acute fractures are defined as being less than 3 weeks old. Langhoff and Anderson showed that after a 4-week delay, the percentage of bony union with cast treatment decreases markedly. Delayed union is defined as failure to heal by 4 to 6 months. Ununited fractures older than 6 months are considered nonunions. These are, of course, arbitrary definitions.
Location
The fracture can also be defined anatomically as distal third (pole), middle third (waist), or proximal third (pole). Waist fractures are the most common, accounting for approximately 80% of all scaphoid fractures. Proximal pole and distal pole fractures are seen in 15% and 5% of cases, respectively.
This classification has prognostic implications. Proximal pole fractures are described as having a poorer rate of healing than more distal fractures, presumably owing to interruption of the blood supply, which enters the bone at or distal to the middle third. Gelberman and Menon showed that the major blood supply is from branches of the radial artery that enter the bone at or distal to the waist at the dorsal ridge. This blood supply accounts for 70% to 80% of the total intraosseous vascularity and 100% of the proximal pole ( Fig. 41.5 ). The osteonecrosis rate for proximal pole fractures approaches 100%.
Orientation
Russe (see Fig. 41.3 ) and later Herbert and Fisher (see Fig. 41.4 ) suggested that the plane of the fracture is important and described horizontal oblique, vertical oblique, and transverse fractures; they concluded that vertically oriented fractures are less stable and, consequently, less likely to heal.
Displacement
Cooney and colleagues and Weber emphasized fracture stability as defined by displacement. Fractures with more than 1-mm stepoff on any view, a scapholunate angle of more than 60 degrees, a lunocapitate angle greater than 15 degrees, or a lateral intrascaphoid angle of more than 20 degrees are considered displaced. Displacement has been shown to affect healing dramatically in conservatively treated scaphoid waist fractures, with nonunion rates as high as 92% being reported.
Comminution
Comminuted fractures are inherently unstable.
Associated Injuries
Scaphoid fractures commonly accompany perilunate dislocations. These unstable injuries require open reduction and internal fixation (ORIF). The specifics of management are discussed in the carpal dislocations section.
Approximately 5% of distal radius fractures are associated with fracture of the scaphoid. These high-energy injuries usually require surgery.
Patient Factors
Smoking has a deleterious effect on the outcome of surgery for scaphoid nonunions. Although there are no conclusive data that smoking affects the healing of acute scaphoid fractures, we nevertheless advise patients to discontinue smoking. Depending on his or her perspective, the surgeon may be more or less inclined to operate on an acute scaphoid fracture in a smoker. The potential economic impact of cast immobilization on the individual is also a factor and may be the principal determinant for surgery in some instances.
We agree that displacement and angulation are more important indicators of prognosis than is the plane of the fracture. However, plain radiographs may not be sufficient to determine whether a scaphoid fracture is truly nondisplaced. Some researchers recommend obtaining a scaphoid CT for all waist fractures, citing the high incidence of nonunion, malunion, and avascular necrosis (AVN) associated with fracture displacement. Their rationale is that the treating physician is obligated to prove that the scaphoid fracture is truly nondisplaced before recommending nonoperative treatment.
In summary, poor prognostic factors for successful nonoperative treatment of scaphoid fractures include late diagnosis, proximal location, displacement or angulation, and possibly obliquity of the fracture line. Scaphoid fractures associated with perilunate dislocations are unstable and require internal fixation.
Management
Disorder or Injury
Treatment clearly remains the most controversial aspect regarding scaphoid fractures. The advent of percutaneous methods of stabilizing the fractured scaphoid, combined with early mobilization, has made surgical management more appealing to patients and some surgeons. However, the concept of surgical treatment of even nondisplaced scaphoid fractures is not new. In 1954 McLaughlin advocated open treatment of all scaphoid waist fractures, acknowledging “the almost universal refusal of surgeons who advocate prolonged immobilization to submit their own fractured naviculars to the ‘long enough’ requisite of this plan of treatment.” The treating physician must weigh these issues with the fact that the vast majority of nondisplaced scaphoid fractures will heal without surgery.
Emergent Treatment
The majority of scaphoid fractures can be treated electively. As noted, immobilization initiated within the first 4 weeks appears to result in satisfactory outcomes for nondisplaced scaphoid fractures. The most common indications for emergent treatment include scaphoid fractures associated with perilunate dislocations (transscaphoid perilunate dislocation) and those associated with dense or worsening median neuropathy. Other indications for emergent treatment include open scaphoid fractures (exceedingly rare) and those associated with vascular injury.
Indications for Definitive Care
Given the potential difficulty in obtaining union, all scaphoid fractures require treatment with some form of immobilization, fixation, or both. Distal pole fractures are generally treated nonoperatively. Proximal pole fractures, comminuted fractures, and scaphoid fractures with associated injuries are generally treated operatively, even if nondisplaced. Treatment of nondisplaced waist fractures remains controversial.
Nonoperative Treatment
Nondisplaced Fractures
Cast immobilization remains the mainstay of treatment for nondisplaced scaphoid waist fractures. The particular type of cast, however, has been the subject of much debate. Almost every wrist position has been advocated, including flexion, extension, radial deviation, ulnar deviation, neutral, and various combinations. Most investigators have recommended including the thumb, but others have included the thumb, index, and middle fingers (three-digit cast). Yet others believe that a simple short arm cast is sufficient.
Undoubtedly, the most controversial aspect of cast immobilization has been the long arm versus short arm debate. The most frequently cited study in this regard was performed by Gellman and colleagues, who concluded that the time to union was faster (9.5 vs. 12.7 weeks; P <0.05) and the nonunion rate lower (0 vs. 8.7%, not significant) when a long arm thumb spica cast was used for the first 6 weeks of treatment. However, a recent cadaver study demonstrated only 0.2 mm of scaphoid fracture motion throughout full forearm rotation in a short arm thumb spica cast. It has been theorized that long arm casts can be detrimental to scaphoid healing because they prevent normal forearm rotation and potentially increase carpal rotation as the patient attempts to use the hand.
Part of this diversity in cast immobilization treatment may be because of the consistently successful results (94% to 98.5%) reported by several investigators with varying cast treatment of fresh nondisplaced fractures. It appears that the exact type of cast is not a critical factor in successful treatment. Consequently, we favor a well-fitting short arm thumb spica fiberglass cast with the wrist in neutral position. The cast is usually changed at 2-week intervals to ensure that it fits snugly. PA, lateral, and ulnar deviation PA radiographs are performed at 6 weeks. If the radiographs are equivocal for healing, a short arm thumb spica cast is reapplied, and a scaphoid CT is obtained. If radiographs do not show that the fracture is healed, regardless of the absence of pain or tenderness, a below-elbow cast is applied for 4 to 6 more weeks. If after a total of 12 weeks of immobilization, radiographic examination fails to show unequivocal healing, a CT scan is performed.
Some evidence indicates that cast immobilization accompanied by electrical stimulation may promote union of un-united scaphoid fractures. However, recent literature suggests that this modality is not effective in hastening union of acute scaphoid fractures. Our experience with electrical stimulation in this setting is anecdotal.
Surgical Treatment
Nondisplaced Fractures
Although the majority of surgeons remain satisfied with cast treatment for nondisplaced scaphoid fractures, there appears to be a growing interest in internal fixation of these fractures. Herbert and Fisher reported a 50% failure rate with nonoperative treatment and suggested that early internal fixation was appropriate for many patients, especially young manual laborers and professional athletes who would not tolerate prolonged cast immobilization. This position was supported by Rettig and Kollias, whose patients returned to athletics within 6 weeks after scaphoid fixation through a volar approach. More recently, percutaneous scaphoid fixation using a cannulated screw through either a dorsal or volar approach has gained popularity. Union rates of 100% have been reported in several series. Furthermore, return to function may be hastened. In one series of percutaneous scaphoid fixation in military personnel, patients achieved union 5 weeks earlier and returned to full duty 1 month sooner than with cast treatment ; similar results were subsequently demonstrated in the civilian population. Whether such an approach is considered overtreatment is controversial. Two meta-analyses of randomized controlled studies on the subject demonstrated earlier time to union in the surgical group but conflicting results on the frequency of union and complications. Although the early studies reported virtually no complications from percutaneous scaphoid fixation, one study demonstrated a sobering complication rate of 29%, including nonunion, hardware problems, and postoperative fracture of the proximal pole of the scaphoid. A more recent systematic review of percutaneous screw fixation demonstrated earlier return to work, earlier radiographic union, and comparable complication rates. Given the variation in the available studies, the decision for early percutaneous fixation is reasonable but the risks of surgery must be understood.
In considering the surgical approaches for percutaneous scaphoid fixation, potential advantages of the volar approach include (1) higher surgeon level of comfort because the traditional open approach for waist fractures is volar, (2) easier ability to obtain proper radiographs because the obligatory wrist extension places the long axis of the scaphoid parallel to the radius, (3) one-step placement of the guidewire, and (4) potentially less likelihood of displacing the fracture. Potential disadvantages of the volar approach include injury to cutaneous nerves and the scaphotrapezial joint and possibly biomechanically inferior fixation because it is more difficult to place the screw tip in the center of the proximal pole. The major advantage of the dorsal approach is that it is easier to center the screw in the proximal pole. This is critical for proximal pole fractures, which should be managed through a dorsal approach. Furthermore, if an arthroscopically assisted approach is chosen, a dorsal approach to fixation is generally preferred. Potential drawbacks of the dorsal approach include (1) being more technically demanding than the volar approach, (2) injury to the extensor tendons, (3) damage to the articular surface of the proximal pole of the scaphoid, and damage to the radius if the screw is prominent, and (4) fracture displacement resulting from the acute wrist flexion required to place the guidewire. In instances in which percutaneous scaphoid fixation is chosen, the authors favor a volar approach for waist fractures and a mini–open approach for proximal pole fractures to protect the extensor tendons and confirm proper seating of the screw.
Percutaneous Scaphoid Fixation
Volar Percutaneous Scaphoid Fixation Technique ( Fig. 41.6 ).
We prefer to wrap the wrist with sterile Coban to maintain the wrist in extension. In this way, the guidewire will be inserted parallel to the radius and approximately 45 degrees toward the ulna. A 0.045-inch guidewire is advanced percutaneously from the radial aspect of the distal pole of the scaphoid into the center of the proximal pole. This guidewire should be perpendicular to the fracture. Proper pin placement is then confirmed fluoroscopically on multiple views. An antirotation pin is then placed parallel to the first pin. A 1-cm incision is made at the guidewire entry site, and a second guidewire is advanced to the distal pole by hand, parallel to the guidewire, to measure proper screw length. After measurement, the cannulated drill is advanced by hand under fluoroscopic control. An Acutrak (Acumed, Beaverton, OR) screw or similar headless screw measuring 2.5 mm shorter than the measured length is then advanced. Maintenance of the reduction is confirmed fluoroscopically, as is proper screw length. The guidewire is withdrawn before final seating. The skin is approximated with a single 5-0 nylon suture, and a sterile dressing and radial gutter splint are applied. At the 2-week follow-up, the suture is removed, and a removable splint is applied. Wrist motion is encouraged, although strengthening is not. Radiographs are obtained at 6 weeks after surgery and then monthly if necessary. Unrestricted activity is permitted once the scaphoid fracture is bridged.
Dorsal Percutaneous Scaphoid Fixation.
The following has been adapted from Slade and colleagues ( Fig. 41.7 ). The wrist is flexed 45 degrees and pronated until the proximal and distal scaphoid poles overlap radiographically. The entry site is determined, and a small incision is made directly overlying it. The extensor tendons are protected, and a small capsulotomy is made. For proximal pole fractures, proper reduction is confirmed. A 0.045-inch guidewire is placed in the center of the overlapping scaphoid rings and is advanced through the volar skin adjacent to the trapezium. The wire is then withdrawn distally to obtain high-quality radiographs by extending the wrist. After proper placement and reduction are confirmed, the wrist is flexed and the guidewire is advanced dorsally and then withdrawn so that the tip is at the level of the subchondral plate of the distal scaphoid pole. The screw length is then measured, and a screw 4 mm shorter than the measured length is selected to maintain the screw within the bone. The guidewire is then readvanced volarly so that it may be retrieved in the event of breakage. An antirotation pin may be necessary. The guidewire is then overdrilled from proximal to distal by hand with the cannulated drill. Proper depth of the drill is confirmed fluoroscopically, the drill is removed, and the screw is advanced over the guidewire until the trailing end is under the articular cartilage. The remainder of the procedure and postoperative care are the same as for the volar approach.
Displaced or Unstable Fractures
The management of acute, displaced scaphoid fractures is far less controversial. These fractures require surgical treatment. Options include closed reduction and percutaneous pin or screw fixation, arthroscopically assisted pin or screw fixation, and open reduction with either pin or screw fixation. Our preference is ORIF with a cannulated screw. In general, the screw will be inserted from the smaller fragment into the larger fragment. The dorsal approach is reserved for proximal pole fractures. The volar approach is safer for waist or distal one-third fractures because the primary blood supply to the scaphoid is dorsal. Bone graft, when necessary, can usually be obtained from the distal radius.
Open Volar Approach ( Fig. 41.8 and see )
A rolled towel under the wrist will facilitate exposure. A 4- to 5-cm zigzag volar incision is made directly over the flexor carpi radialis (FCR) tendon ending distally at the tuberosity of the scaphoid. One should be careful in dividing the skin to not injure the palmar cutaneous branch of the median nerve, which often lies in the skin flap adjacent to FCR tendon. The FCR tendon is mobilized and retracted ulnarward, and the radial artery is retracted radialward. The small crossing superficial branch of the radial artery is identified and is usually ligated and divided. The posterior wall of the FCR sheath is divided longitudinally, and the underlying pericapsular fat is exposed. This too is divided, and the multiple vessels in this layer are cauterized with a bipolar instrument. The scaphoid tuberosity is exposed, and dissection is carried proximally through the volar capsular ligaments, incising only as much as is necessary to adequately visualize the fracture site. Depending on the interval from injury, early callus may need to be curetted from the fracture. Kirschner wires (K-wires) may be placed as joysticks, with care to avoid the projected path of the screw. Typically, the distal pole rests in flexion and pronation, and the proximal pole rests in extension. The fracture is then reduced and stabilized provisionally with a distal-to-proximal axial K-wire, radial to the projected path of the screw. Proper reduction and pin placement are confirmed fluoroscopically. The volar lip of the trapezium adjacent to the scaphotrapeziotrapezoid (STT) joint is removed with a rongeur to facilitate guidewire placement. In most instances, we will supplement the fixation with a headless screw inserted from the distal pole of the scaphoid.
With the Acutrak system, a 0.045-inch guidewire is then inserted in a distal to proximal direction. After the proper angle of insertion is confirmed fluoroscopically, the guidewire is advanced into the subchondral plate of the proximal pole. Proper guidewire placement must be confirmed fluoroscopically in multiple planes. Ideally, the guidewire should traverse the central third of the proximal pole. The proper screw length is then measured over the guidewire, with care to ensure that the depth gauge is contacting the scaphoid. In general, we select a screw that is 2.5 mm shorter than the measured length for several reasons: the scaphotrapezial joint is oblique to the screw insertion angle, which can result in screw impingement on the trapezium; the leading end of the screw must not be allowed to penetrate into the radioscaphoid joint; and the Acutrak screw is tapered and cannot be withdrawn without compromising the fixation. Before the guidewire is overdrilled, a 0.045-inch antirotation pin is inserted across the fracture. The guidewire is then overdrilled by hand under fluoroscopic control until the tip of the drill reaches the anticipated final screw position. Failure to drill completely may result in distraction of the fracture as the screw is advanced. The screw is then inserted over the guidepin while the surgeon inspects the fracture site for rotation or gapping. The guidepin is withdrawn before final seating of the screw to prevent incarceration. Stability is checked by moving the wrist and checking the fracture site, and final radiographs are obtained. If there is any concern about fracture stability, the antirotation pin is left in place and cut beneath the skin. Proper reduction and hardware placement is confirmed fluoroscopically. The volar ligaments and capsule are closed with absorbable 2-0 suture in horizontal mattress fashion. The rest of the closure is routine. A sterile dressing and splint are applied.
Open Dorsal Approach
The dorsal approach (see ) is preferred for proximal pole fractures and scaphoid fractures associated with perilunate dislocations. In these instances, a 3- to 4-cm dorsal longitudinal incision is made over the wrist. If the injury is limited to the scaphoid, the fourth compartment extensor retinaculum is opened distal to the radius. A ligament-preserving capsulotomy is made. For transscaphoid perilunate injuries or cases in which dorsal distal radius bone graft will be obtained, a longer incision is required. In those instances, the retinaculum is then incised over the extensor pollicis longus (EPL) tendon, the Lister tubercle is osteotomized, and a longitudinal capsulotomy is performed deep to the fourth compartment. Care is taken to avoid dissection at the dorsal ridge to minimize iatrogenic vascular injury to the scaphoid. The fracture is reduced using K-wires as joysticks if necessary. A provisional K-wire is placed across the fracture site; this will also serve as an antirotation pin. Proper reduction and pin placement are confirmed fluoroscopically. For small proximal pole fractures, we prefer to use the mini-Acutrak screw. The starting point for the guidewire is adjacent to the membranous scapholunate ligament. The wrist must be acutely flexed to ensure proper wire placement. After proper guidewire position is confirmed fluoroscopically on multiple views, proper screw length is measured. The guidewire is then advanced distally so that it may be retrieved in the event of inadvertent wire breakage. The screw selected is 2 to 4 mm shorter than the measured length. As with the volar approach, complete overdrilling is necessary to minimize the likelihood of fracture gapping. Care must be taken to ensure that the trailing end of the screw is seated beneath the articular cartilage. Occasionally, the proximal pole may be too small for the mini-Acutrak screw. In such instances, a countersunk 2.0- or 2.7-mm AO screw can be of value. After final radiographs have been obtained, the pins are removed, and the capsulotomy is repaired with 2-0 Vicryl sutures. If the EPL has been mobilized, the extensor retinaculum is approximated deep to the EPL, which is left in the subcutaneous tissue. The subcutaneous tissue is approximated with 3-0 Vicryl and the skin with 4-0 Monocryl running subcuticular suture. Steri-Strips, a sterile dressing, and a splint are applied.
Pitfalls and Avoidance of Complications
In practice, the principal complications are primarily related to screw placement. These can be minimized by strict attention to the details of proper guidewire positioning. With the volar approach, a common error is to place the guidewire too anteriorly in the scaphoid tuberosity, which can result in fracture of the tuberosity or breakout at the volar waist of the scaphoid. This complication can be averted by placement of a rolled towel under the wrist to provide wrist extension and removal of a small amount of the volar-radial corner of the trapezium with a rongeur. This permits the guidewire to be inserted more centrally in the distal pole. Placement of the guidewire off axis may result in displacement of the fracture as the screw threads push against the endosteal cortex. This may also occur if the fracture line is not perpendicular to the long axis of the scaphoid. For oblique fractures, the guidewire may be adjusted to approximate the perpendicular angle, and a second pin will help to resist displacement when the screw is tightened. Gapping of the fracture may occur if the path is not completely predrilled. Inadvertent penetration into the radiocarpal joint may occur in an area that we have termed the “blind spot” dorsoradially. Because the scaphoid is triangular in cross section, a long screw may appear to be in the bone on both the PA and lateral views. A 50-degree hyperpronation PA view will place this face into relief and potentially identify otherwise unrecognized screw penetration.
With the dorsal approach, failure to fully seat the trailing end of the screw will eventually damage the articular surface of the radius. For this reason, we routinely select a screw that is 4 mm shorter than measured when placing dorsally. Fracture of the proximal pole may occur if the guidewire is not overdrilled properly.
Postoperative Care and Rehabilitation
The sutures are removed, and a short arm thumb spica cast is applied at the first postoperative visit. A baseline scaphoid series is obtained out of plaster. The cast is changed at 2- to 3-week intervals. Radiographs are obtained at 6 weeks after surgery. If union is apparent, the cast is discontinued. For equivocal cases, the cast is reapplied, and a scaphoid CT is obtained 2 weeks later. When the CT demonstrates greater than 70% union, the cast is discontinued, and occupational therapy is initiated, focusing on restoration of wrist motion before any strengthening.
Complications
Surgical
Wrist stiffness and hypertrophic scarring are potential complications related to surgery. With the volar approach to the scaphoid, several cutaneous nerves are at risk, including the terminal branches of the lateral antebrachial cutaneous nerve, the radial sensory nerve, and the palmar cutaneous branch of the median nerve. These are best avoided by making an incision directly over the FCR tendon. The superficial branch of the radial artery courses from proximal-radial to distal-ulnar directly over the FCR and may be ligated and divided if injured or needed for exposure. Excessive division of the volar extrinsic ligaments (radioscaphocapitate and long radiolunate) must be avoided because these are difficult to repair, and disruption could potentially lead to ulnar translocation of the carpus.
On the dorsal side, cutaneous nerves at risk include the dorsal branch of the radial sensory nerve and the communicating branch between the dorsal sensory radial and ulnar nerves. The digital extensors are at risk with the percutaneous approach, and the obliquely coursing EPL is at risk with the open approach. Care must be taken not to disrupt the blood supply to the scaphoid, which enters the dorsal ridge from a branch of the radial artery, or the scapholunate ligament.
Complications of screw fixation include screw tip penetration, inadequate seating of the trailing end deep to the articular cartilage, fracture displacement (if the screw is placed off axis or out of plane of the fracture), and fracture at the entry site of the screw. Thorough knowledge of the technical details of the particular screw to be used will mitigate these risks.
Delayed Union
The location of the scaphoid fracture has a significant impact on time to union. The average healing time ranges from 4 to 6 weeks for tuberosity fractures, 10 to 12 weeks for waist fractures, and 12 to 20 weeks for proximal pole fractures. Therefore the generally accepted upper limit of “normal” healing time of 4 months must be considered in the context of fracture location.
Some investigators have suggested that noninvasive electrical stimulation be applied for failed treatment. After a review of the experience with electrical stimulation, Osterman and Mikulics concluded that noninvasive methods of electrical stimulation are most appropriately applied to patients who have failed previous bone grafting, have well-aligned waist fractures, do not have significant collapse patterns, and do not have a small proximal fragment or significant arthritis. In the absence of controlled studies showing the efficacy of electrical stimulation, we prefer bone grafting as the next step after failure of cast treatment.
Nonunion
Operative Treatment
The decision whether to operate on a scaphoid nonunion is relatively straightforward when the patient is symptomatic. Not infrequently, however, patients come in with new-onset wrist pain after an injury, and radiographs demonstrate a long-standing nonunion. The physician could simply advise the patient to splint the wrist, anticipating that the pain will resolve and the patient would return to the asymptomatic un-united state. This is not necessarily true. The issue of advising surgery for the patient with an asymptomatic nonunion is a difficult one. The experience of Mack and colleagues, Ruby and colleagues, and Lindstrom and Nystrom showed that scaphoid nonunion results in wrist malalignment and arthritis if left untreated for 5 to 10 years. If surgical treatment of the fracture is selected, the goals of obtaining good alignment of the wrist and scaphoid as well as union should be kept in mind. The techniques of operative treatment of the fracture presently include internal fixation, bone grafting, or both. Current salvage techniques include radial styloidectomy, distal pole scaphoid resection, proximal row carpectomy, either complete or partial arthrodesis, and combinations.
Operative Treatment (see Through )
The surgical approach depends on a number of factors, including the following:
- 1.
Fracture location: In general, the volar approach is used for waist and distal pole fractures, thereby (a) protecting the principal blood supply to the scaphoid, which enters dorsally ; (b) allowing for good visualization of the nonunion site; (c) facilitating deformity correction; and (d) facilitating hardware placement from the smaller into the larger fragment. The dorsal approach is reserved for proximal pole fractures.
- 2.
Deformity: Scaphoid deformity cannot be reliably corrected through a dorsal approach.
- 3.
Avascularity of the proximal pole: This most commonly involves fractures of the proximal pole. In these cases, the approach will be based on whether a vascularized graft will be used and what the anticipated source of the graft is (see Vascularized Bone Grafts ). In the absence of collapse of the proximal pole caused by AVN, the scaphoid alignment is usually well maintained and the approach is dorsal.
Bone Grafting
Bone grafting without supplemental hardware is the oldest technique for treating established nonunion or delayed union. The autogenous bone graft serves several purposes. It is osteoconductive and osteoinductive, and it provides a source of osteoprogenitor cells. In addition, it can be used as a structural graft, contoured to fit existing defects or to correct three-dimensional deformity. Donor sites include the iliac crest, distal radius, and proximal ulna. Histomorphometric studies have demonstrated that the quality of cancellous graft is better from the iliac crest than from other sources. An admittedly flawed clinical study by Hull and colleagues confirmed that grafts from the iliac crest were superior to those from the distal radius in the management of scaphoid nonunions. A prospective, randomized study comparing distal radius versus iliac crest grafts for scaphoid nonunion concluded that these donor sites were equivalent. Our preference is to use iliac crest grafts when significant deformity correction is necessary and distal radius graft for well-aligned nonunions and those associated with AVN.
Several different methods of bone grafting have been described over the years. Corticocancellous or cancellous interposition grafts and anterior wedge grafts are the most commonly used at present. The original Matti technique as described in 1937 consisted of excavation of the proximal and distal fracture fragments through a dorsal approach and placement of cancellous graft within these two cavities to act as an internal fixation device as well as a focus for osteogenesis ( Fig. 41.9 ). In 1960 Russe described a volar approach also using cancellous graft, and later he advocated a doubled corticocancellous graft ( Fig. 41.10 ). Because the blood supply of the scaphoid is primarily dorsal, the volar approach is the most popular today. The Matti-Russe–type graft is indicated when the nonunion is not associated with significant dorsal intercalated segment instability (DISI) of the wrist. When DISI is present, an anterior wedge graft after the method of Fisk, Fernandez, or Cooney and associates has been the preferred option, although a nonstructural graft may be used. Green pointed out that the Matti-Russe technique has a lower success rate when the proximal pole is avascular, as documented by no bleeding of the bone at surgery. In his experience, he had a failure rate of 100% in five patients in whom there was no bleeding of the proximal pole at the time of surgery. He and others have shown that radiographic avascularity is not a reliable indicator of actual vascularity. Our experience with the Matti-Russe method is consistent with that of most investigators, and 80% to 90% healing rates can be expected.
Vascularized Bone Grafts.
Over the past two decades, increasing attention has been paid to the use of vascularized bone grafts in the management of difficult nonunions. These have included the volar pronator pedicle graft ; the dorsal Zaidemberg 1,2 intercompartmental artery pedicle graft ( Fig. 41.11 ); the Fernandez vascular bundle implant ; the free vascularized iliac crest graft ; and, most recently, the free vascularized medial femoral condyle graft. In general, the results using these techniques have been equivalent to those of more conventional nonvascularized methods. A study from the Mayo Clinic concluded that the dorsal-based Zaidemberg graft may not be capable of predictably correcting humpback deformities, which may be better managed with a nonvascularized corticocancellous or a vascularized medial femoral condyle graft through a volar approach. We believe that the structural requirements of the graft are more important in graft selection than its vascularity. There clearly does appear to be a role for the dorsal pedicle grafts in the management of well-aligned, proximal pole nonunions associated with AVN. The vascularized medial femoral condyle graft is reserved for the disastrous scaphoid nonunion associated with carpal collapse and AVN or failure of prior surgery.
Modified Matti-Russe Technique (see ).
The volar approach for scaphoid nonunion is the same as for the acute fracture (see Open Volar Approach and ). Only as much of the volar capsular ligament complex is incised as necessary to adequately visualize the nonunion site. At this point, the radioscaphoid joint and the majority of the volar surface of the scaphoid are exposed. The fracture site is usually obvious, but on occasion, a wrinkle in the articular cartilage of the scaphoid may be the only clue. Fingertrap traction and wrist extension will optimize the exposure. An opening is made in the volar nonarticular cortex of the scaphoid at the level of the fracture. The opposing cavities of the two fragments are then excavated. Russe advocated using only hand instruments for this step, but we use the Midas Rex (Midas Rex Pneumatic Tools, Ft. Worth, TX) with an X2 bit for the preliminary curettage and then complete the excavation with straight and curved handheld curettes. If the proximal pole is very small and avascular, the curettage is carried to the subchondral bone level. Because the articular cartilage is nourished by synovial fluid, the proximal pole is converted into an osteocartilaginous graft. Particular care must be taken not to penetrate the proximal pole if this maneuver is used. After adequate curettage (bleeding cancellous bone is identified) of both fragments, cancellous autograft is harvested from either the iliac crest or distal radius. For the distal radius harvest, the skin incision is extended proximally along the FCR tendon, which is then retracted, and the distal edge of the pronator quadratus is elevated to expose the metaphysis. Either corticocancellous matchsticks or cancellous chips are then harvested and packed tightly into the defect by overdistracting the fragments. K-wires may be used as “joysticks” for this maneuver. Stability is then checked by manipulation of the wrist in radial-ulnar deviation and flexion-extension. In most instances, we supplement the fixation with a headless screw inserted from the distal pole of the scaphoid. Proper reduction and hardware placement are confirmed fluoroscopically. Closure is the same as for the acute fracture. After surgery, a radial gutter splint is used for 7 to 10 days, after which a solid, short arm thumb spica cast is applied. The cast is changed every 2 to 3 weeks. At 6 weeks after surgery, radiographs are taken out of plaster. A short arm thumb spica cast is continued for 6 more weeks or until radiographic results indicate that healing is complete. We almost always obtain a CT scan of the scaphoid to confirm adequate healing.
When angulation is present at the fracture site with concomitant DISI, the Fisk-Fernandez method is most appropriate in our hands.
Fisk-Fernandez Technique ( Fig. 41.12 ).
According to Fernandez and Cooney and associates, it is helpful to measure the normal scaphoid to determine the amount of bone to be resected and the size and shape of the bone graft. A volar wrist approach similar to that in the Matti-Russe technique is used. Although volar angulation at the fracture site makes visualization difficult, wrist extension and ulnar deviation reveal the fracture. To reduce the DISI deformity of the carpus, a small laminar spreader is placed into the scaphoid fracture site and opened as much as necessary. This maneuver usually reduces the fracture and the carpal instability pattern. Alternatively, one may place 0.054-inch K-wires in the proximal and distal fragments as joysticks and manipulate them into the corrected position. When the carpal collapse is severe, the lunate is first reduced with respect to the radius by palmar translation through the midcarpal joint. A 0.062-inch K-wire is then driven from the radius into the lunate as temporary fixation. This maneuver will result in a gap between the two scaphoid fragments that approximates the size of the necessary graft.
The fracture site is then curetted sufficiently to expose good, bleeding cancellous bone on both surfaces. A corticocancellous wedge-shaped bone graft is harvested from the iliac crest. In some patients, the entire scaphoid is shortened, and there is volar and ulnar angulation at the fracture site. This may be due to comminution dorsally at the time of fracture or erosion at the fracture site owing to motion between the fragments. In this situation, after reduction is obtained, there will be a dorsal–ulnar gap as well as a larger volar defect. Therefore a trapezoid-shaped bicortical graft is trimmed to fit snugly into the defect with the cortical surface anterior. The graft is then placed followed by a headless screw inserted from distal to proximal, transfixing the graft. If the proximal fragment is very small or has to be extensively curetted because of avascularity, we will use K-wires and advance them into the lunate to improve fixation. An alternative method for proximal pole fractures is to use a dorsal approach with bone graft and either K-wire or screw fixation. Stability is checked by observing the fracture site while manipulating the wrist. Radiographs are taken in the anterior-posterior (AP) and lateral planes to check alignment and pin placement. A radial gutter is applied for 7 to 10 days. This is then changed to a short arm thumb spica cast with the wrist in neutral position for 6 weeks; the cast is changed every 2 to 3 weeks. Radiographs and, if necessary, a CT scan, are obtained. As indicated, casting is continued for 6 to 9 more weeks until solid union is established by radiographs. This cast is also changed every 2 to 3 weeks.
Internal Fixation.
The indications for internal fixation have been mentioned and include displaced acute fractures, fractures associated with ligamentous injury such as transscaphoid perilunate fracture-dislocations, delayed union, and nonunion when bone grafting alone is insufficient to provide adequate stability. A relative indication is an inability to tolerate long-term cast immobilization. The internal fixation techniques described include K-wires, screws, plates, and staples, although the latter have not achieved much popularity and are not in general use. Of these, K-wires are the simplest, and screws are the most secure.
For many years, the Herbert screw system was one of the most popular methods of scaphoid fixation. This ingenious headless device consists of a smooth shaft with threads at both ends with differing pitch. As the screw is advanced, compression is obtained across the fracture site by putting greater pitch in the leading end threads than in the trailing end threads, leaving the entire screw within the bone. This solid screw could be inserted either freehand or through an alignment-compression jig. Compared with K-wires, however, the technique is demanding and unforgiving.
A major advance in scaphoid fixation has been the development of small, cannulated screws. With these, the surgeon has the ability to position a small guidewire optimally with minimal trauma to the scaphoid before committing to the screw position. Over the past decade, the most popular scaphoid screws have been the Acutrak screw and the AO 3.0-mm cannulated screw. These two screws generate compression by different methods. The AO screw is a headed, partially threaded screw, which generates compression through the standard “lag” effect. A threaded washer can be used to reinforce the near cortex purchase. The Acutrak screw is a headless, tapered, cannulated screw with a continuously differential pitch that generates compression as the screw is advanced. It should be noted that, unlike the AO screw, the compression generated with the Acutrak screw is dictated by the length and size of the screw and the position of the fracture. Longer, standard screws placed with the trailing end (entry site) nearest to the fracture will generate more compression than shorter, mini-Acutrak screws with the fracture nearest the leading end of the screw.
Several studies have compared various screws used in fixation of the scaphoid. Shaw demonstrated that the AO 3.5-mm cannulated screw generates 2.5 times more compression than the Herbert screw. Newport and colleagues concluded that the Howmedica Universal Compression Screw was superior to its competitors. This finding was not supported by Toby and associates, who identified iatrogenic fractures at the screw insertion site where the larger universal compression screw was used. In that study, the Acutrak, Herbert-Whipple, and AO screws were superior to the Herbert screw.
In clinical practice, Trumble and colleagues found no difference between the Herbert screw and the cannulated AO screw for scaphoid nonunions. However, they did observe that the location of the screw in the proximal scaphoid was important. Time to union was significantly shorter when the screw was positioned in the central third of the proximal scaphoid. In a biomechanical study of Acutrak screw length, Dodds and colleagues demonstrated less fracture motion with longer screws and concluded that the ideal screw length should be about 4 mm less than the scaphoid length. However, another biomechanical study showed that a screw perpendicular to an oblique fracture was just as strong as a centrally based screw although the length may be shorter. In summary, attention to technical details such as screw length and position is critical to ensure optimal fracture stability.
Special Situations
Proximal Pole Fractures.
As noted earlier, vascular studies by Gelberman and Menon and Taleisnik and Kelly established that fractures through the proximal third of the scaphoid have a high likelihood of devascularizing the proximal fragment because most of the blood supply enters the bone distal to this level. The proximal pole is covered almost entirely with cartilage and has a poor blood supply apart from its connection to the rest of the scaphoid. With nonoperative treatment, proximal pole fractures may take as long as 20 weeks to heal and, even then, the nonunion rate is estimated to be 34%. Consequently, many surgeons, us included, favor internal fixation for even nondisplaced proximal pole fractures. We approach these fractures through a small dorsal incision as described earlier. For nonunion of proximal pole fractures, we recommend a cancellous bone graft and internal fixation through a dorsal approach after the method of DeMaagd and Engber. It is particularly important in such areas to perform a thorough and aggressive curettage of the proximal pole to bleeding cancellous bone if present or subchondral bone if not. This is followed by tight packing with cancellous graft (Matti technique) and internal fixation with a screw or K-wires. The same surgical treatment is recommended for AVN of the proximal pole if there is no collapse or significant arthritis despite limited success in others’ experience. Several investigators have recommended vascularized bone grafts for these small proximal pole fractures when the blood supply is marginal or absent. Zaidemberg and colleagues reported on 11 patients who were treated with vascularized bone graft from the dorsal radial aspect of the distal radius based on the ascending irrigating branch of the radial artery in the snuffbox. They had a 100% union rate determined by radiography, and five of the patients had previously failed Matti-Russe procedures. However, these outstanding results have not been replicated by other authors, who reported union rates of 72% to 75%. This procedure is not indicated for AVN associated with carpal collapse.
In the setting of collapse of the proximal pole, significant arthritis, or a failed bone graft, several options are available. Revision surgery in this setting produces modest results at best. Excision of the proximal fragment with or without replacement with a tendon or free osteochondral graft are scaphoid-preserving options. We have been reluctant to perform proximal pole excision alone based on concern for scapholunate dissociation owing to the necessary loss of the stabilizing effect of the scapholunate ligament. Watson and Ballet suggested excision of the entire scaphoid and capitate–hamate–triquetrum–lunate (“four-corner”) arthrodesis. Proximal row carpectomy and limited or complete wrist arthrodesis are reasonable salvage procedures for failed bone graft of a proximal pole nonunion.
Distal Pole Fractures.
The most common distal pole fracture is an extraarticular avulsion injury of the tuberosity. The small fragment is best visualized on a semipronated view ( Fig. 41.13 ). Cast treatment for several weeks is usually sufficient. In cases of symptomatic nonunion, this fragment may be excised. Distal one-third fractures of the body of the scaphoid, if recent and nondisplaced, should heal in 4 to 8 weeks with cast treatment. A less common, often subtle, fracture is the distal pole vertical intraarticular fracture, which may be visible only on polytomography or CT scan. If nondisplaced, these should be treated with a cast; if displaced, ORIF should be performed. Nonunion in this well-vascularized area is uncommon. A 10-year follow-up study of 41 distal pole scaphoid fractures treated nonoperatively demonstrated excellent results, with one asymptomatic nonunion and seven patients with asymptomatic STT arthritis.
Salvage Procedures.
If the decision is made not to attempt to achieve union of the scaphoid, there are several alternatives. It should be emphasized that these procedures are usually indicated after failure of primary treatment, which in most cases will have been an operative attempt to gain union. Given the high rate of success in obtaining union with bone grafting techniques, these methods are necessary in relatively few patients.
You're Reading a Preview
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here