Blood supply

The scaphoid articulates with the radius, trapezium, trapezoid, lunate, and capitate, and about 80% of the surface is covered by hyaline cartilage. The scaphoid is the only carpal bone bridging the proximal and distal carpal rows, and its shape resembles a twisted peanut. The major blood supply to the scaphoid is through the radial artery ( Fig. 8.1 ). The branches of the radial artery enter the scaphoid along the dorsal ridge of the distal scaphoid, which supply 70% to 80% of the scaphoid. The superficial palmar arch gives volar branches entering the volar tubercle at the distal end of the bone and supplies 20% to 30% of the distal portion of the scaphoid ( Fig. 8.1 ). The proximal pole has poor vascularity, which obtains very limited blood supply from interosseous vessels and small vessels in the radioscapholunate ligament.

Classification of scaphoid fracture according to the location of the fracture.

The distal pole, waist, and proximal pole of the scaphoid

A practical classification of the scaphoid is distal pole (tubercle), waist, and proximal pole fracture ( Fig. 8.2 ). Tubercle fractures are less common and usually do not cause clinical consequences. These can be treated for comfort, as with a wrist sprain.

Vascularity of the scaphoid.

(From Chung KC. Operative Techniques of the Hand and Wrist . 4th ed. Elsevier; 2022).

The waist of the scaphoid is relatively nonarticular where blood vessels enter, mostly on the dorsal-radial aspect and at the distal pole. The scaphoid is tethered to the lunate proximally by the scapholunate ligament, distally by the trapezoid and trapezium, and its movement is also constrained by the extrinsic carpal ligaments.

True proximal pole fractures (the entire fracture line is proximal to the capitolunate articulation, i.e., in the proximal 20% of the scaphoid ) are uncommon and hence difficult to study ( Box 8.1 ). It is not clear whether these merit initial management distinct from scaphoid waist fractures.

An oblique fracture may occur, which is difficult to classify. Multiple fragments, a humpback deformity, or remarkable displacements of fragments result in unstable fractures.

Common sites of the fracture, displacements, and consequences

Most scaphoid fractures involve the scaphoid waist ( Box 8.2 ). Nondisplaced fractures may not be visible on radiographs and can be difficult to diagnose. It is possible that a nondisplaced fracture might heal without cast, splint, or screw immobilization unless a reinjury displaces the fracture, but this is difficult to study. Nondisplaced fractures heal with protection and time. It is not clear that cast immobilization is needed if a person limits activity, but the shortest studied period of casting is about four weeks. Screw fixation of a nondisplaced fracture is an option to avoid a cast, but it not necessary for union.

Separation of the scaphoid fracture fragments (fracture displacement) is associated with a risk of nonunion. Nonunion may lead to wrist arthritis that causes variable symptoms. The association of a malunited fracture with arthritis is less clear. The potential for arthritis in a young, active individual might be the reason that this fracture gets so much attention.

Displacement is difficult to diagnose, but CT images made in planes defined by the long axis of the scaphoid may help. Some fractures may be well aligned on imaging, but mobility can be seen on wrist arthroscopy in experimental settings. It is not clear how instability without malalignment affects healing and arthritis.

Clinical presentations

A scaphoid fracture should be suspected when a person falls on their outstretched hand or has sudden forced extension off the hand and develops radial sided wrist pain. Clinical examination may show swelling and tenderness in the anatomical snuffbox. If there is anatomical snuffbox tenderness along with scaphoid tubercle tenderness and a positive scaphoid axial compression test, the patient has a high likelihood of having scaphoid fracture. When radial deviation of the wrist brings the scaphoid to palmar flexion, tenderness at the distal pole (tubercle) palpated just distal to the radial styloid indicates possibility of scaphoid fracture. However, axial compression test may risk fracture displacement; therefore it is not used by many clinicians.

Plain radiographs should be taken immediately if the cause of injury is clear and wrist pain and anatomical snuffbox pain/tenderness are present. Besides routine carpal views, the scaphoid view should be taken. This is with the hand in a fist with the thumb covering the dorsum of the middle phalanges of the index and middle fingers, wrist slightly extended and ulnar deviated, and the forearm pronated ( Fig. 8.3 ).

Radiographic views (A to I) commonly used. It is usual to obtain radiographs in different rotations. The posterior-anterior (PA) (A) and lateral (I) radiographs are orthogonal projections. (B) shows an image taken in PA view. In addition, the semi-supine (C) view gives an assessment of the radioscaphoid joint (D) while the semi-prone view (G) provides a good assessment of the distal scaphoid (H). However, each of these views permit overlap with other carpal bones. The elongated scaphoid view (e.g., Ziter view) taken in ulnar deviation of the wrist (E for a PA view and F for a lateral view) reduces the overlap with neighboring joints.

(From Ziter FM, Jr. A modified view of the carpal navicular. Radiology . 1973).

Factors associated with a higher probability of fracture include pain with ulnar deviation, scaphoid tenderness at the tuberosity or in the anatomical snuffbox, and no distal radius tenderness.

In such a patient with these clinical signs of scaphoid fracture, if initial good quality radiographs with four to five scaphoid views do not show a clear scaphoid fracture, a fracture should still be suspected. In these patients, whose initial radiographs are negative, radiographs are repeated one or two weeks later or a CT scan is ordered.

Occasionally the patient with a scaphoid fracture has a delayed (sub-acute) presentation of symptoms more than 4 weeks after injury. This commonly occurs with a non-displaced waist fracture.

Radiographs

A “scaphoid series” of radiographs includes oblique views and a view made with the wrist in ulnar deviation so that the scaphoid is extended and easier to evaluate ( Fig. 8.3 ). In a high-probability setting, normal radiographs do not exclude fracture. In this circumstance, the injury is referred to as a “suspected scaphoid fracture.” The probability of a scaphoid fracture in a person with radial sided wrist pain and tenderness is somewhere between 7% and 20%, the former representative of the relatively litigious United States and the latter more typical elsewhere. The absence of dorsal soft tissue swelling or an abnormal scaphoid fat stripe on radiographs reduces the likelihood of fracture. Waist fractures proximal to the dorsal scaphoid ridge tend to be less displaced and may be more difficult to detect. In the lateral plain radiographs, interscaphoid angle, i.e., the angle formed by longitudinal axes of the distal and proximal fragments, indicate stability of the fracture, with 35 degrees considering as very unstable. In the lateral view, humpback deformity, i.e., dorsal apex deformity, may be found, and dorsal intercalated segmental instability (DISI) may also present.

Pain in the anatomical snuffbox on ulnar deviation of the wrist within 72 hours of injury and pain over the scaphoid tuberosity at 2 weeks independently predict a fracture. The proportion of patients with a scaphoid fracture confirmed when clinically suspected and initial good quality radiographs reviewed by competent clinicians did not reveal a fracture has been variously reported between 3% and 13%.

Both CT and MRI can have false-positive scans and are not routinely necessary for diagnosis. However, CT can be used as a supplementary method of diagnosis ( Fig. 8.4 ), especially for nondisplaced fractures (see below).

CT scans of two scaphoid fractures with coronal (A and C) and sagittal (B and D) multiplanar reconstructions in the scaphoid plane. (A and B) A clear fracture with an even gap but with no step. (C and D) A tiny step on the radial surface of the scaphoid on the coronal reconstruction, but the sagittal reconstruction shows that the distal scaphoid has bent forward with a step on the palmar surface and a gap opening out on the dorsal surface. (D) shows a mild initial humpback deformity in this acute scaphoid fracture.

Implications of scaphoid fracture displacement

Fracture displacement is the factor most strongly associated with nonunion. Displacement can cause any combination of gap, translation, and angulation. We lack consensus definition and measurement of displacement. A 1-mm gap or translation between fracture fragments is often described as displacement. ^, Others use a threshold of 2 mm.

Radiographs have only a few projections, usually up to five, and therefore could underestimate the magnitude of displacement. Displacement is difficult to measure on radiographs, in part due to the complex anatomy of the scaphoid and the exact path of the x-ray beam.

Good-quality CT scans can give a better assessment of displacement but still depend on the choice of CT slices for the measurement. CT scan is often used in research ( Fig. 8.4 ). To be useful, the sagittal and coronal planes of the CT images are best aligned with the long axis of the scaphoid. Movement of fracture fragments on arthroscopy is occasionally observed in a well-aligned fracture on CT scan, but it is not clear that this is associated with an increased risk of nonunion.

Other measures of scaphoid deformity might be useful for diagnosing displacement, such as angulation of the fracture ( Fig. 8.4 D) or an altered height:length ratio, but these are less well studied.

Diagnosis of fracture union

Confirming scaphoid union is challenging. Conventional radiographs may not confirm union due to the scaphoid’s complex shape. Even with a complete set of radiograph projections, unless the x-ray beam is directed in line with the fracture, the proximal and distal fragments will overlap across the fracture site, giving the appearance of bone bridging across the fracture site. The five views of the scaphoid could suggest uncertainty about union if the fracture site can be identified, there is a persistent gap but only across part of the scaphoid, decrease in lucency around the fracture, or in a small number a clear gap across the scaphoid, suggesting nonunion. Although radiographs can raise concern, they are, on their own, unreliable for diagnosis of union.

CT scans offer another way to assess unions. CT scans help to classify the state of the union early and permit a clinical decision on whether immobilization can be discontinued, modified, or whether a tenuous bony bridge or nonunion needs internal fixation. But diagnosis of union on CT may also not be sufficiently reliable or accurate.

One approach is to arrange a CT scan on all scaphoid fracture patients at around six weeks as the radiation dose is low. It is convenient for the patients as the cast can still be on and convenient for the radiology department as only one imaging modality is used. If a nonunion is confirmed, then immediate surgery can be offered and arranged. If union is confirmed, the cast can be removed, and rehabilitation can commence.

Radiographs or CT both have limited reliability and accuracy for diagnosis of union. That means that imaging alone cannot be used for decisions about duration of cast immobilization or offers of surgery for delayed or nonunion. Consequently, it might be best to limit immobilization to a set amount based on evidence and not revise that for discomfort or imaging appearance. An alternative is to perform a routine CT scan 6 weeks after fracture, but that might be associated with overdiagnosis and overtreatment.

The concept of “partial union” measured on CT scans needs further evidence and consideration. Occasionally, union can occur across only a proportion of the fractured surface. This has been termed “partial union,” and this bridge is weaker than if the whole fractured surface were bridged. On this assumption, the wrist may be protected using a removable splint for around 8 weeks and repeating the CT scan as in most the partial bridge of union slowly consolidates ( Fig. 8.5 ). At present, we do not know if the diagnosis and quantification of partial union are reliable or accurate

A proximal pole fracture involves the proximal 20% of the scaphoid. (A) Shortly after fixation. (B) At 1 year showing no obvious loosening or screw movement, and the appearance suggests union. (C) CT scan at 26 weeks showing union dorsal to the screw in particular. (D) CT scan at 1 year confirming union but also demonstrating that the edges of the screw are prominent. Even after fixation, union can be slow, probably reflecting compromise of vascularity.

Some clinicians permit use of the hand without splintage if there is more than 50% union Fig. 8.5 but protect it from extreme loads, e.g., when there is a risk of falling as in contact sports. If bony bridging is slight (<15%–20%), an alternative to the “splint and review” path could be compressing the fracture site with a screw with or without a bone graft. At present, there is little experimental evidence on the concepts and strategies for “partial union.”

After surgery, diagnosis of union is also hindered by the presence of implants. Loose implants indicate a likely nonunion, but if the implant does not migrate, is it possible that the nonunion will not progress to humpback deformity or arthritis? Implant-related reasons need to be identified, as do correctable reasons such as loosening or infection. Biological reasons such as osteonecrosis need to be noted. We need more evidence in this realm.

Treatment goals

Why treat scaphoid fractures? The goal of scaphoid fracture treatment is union. Nonunion can be associated with a characteristic pattern of wrist arthritis described as scaphoid nonunion advanced collapse (SNAC). Factors associated with the noted variation in radiographic evolution of SNAC, impairment of wrist motion, and levels of discomfort and incapability are not well understood. One study reported very early development of DISI and arthritis. The proportion with arthritis was greater and its progression was quicker in nonunions of mid-waist fractures. Surgery to gain union is less appealing once arthritis is apparent.

Nonunions may also be well-aligned and stable, with fibrocartilage limiting motion of the fragments. The proportion of fractures having such “fibrous nonunion” is unknown, and its natural history has not been reported.

Malunion of a displaced scaphoid fracture leads to carpal malalignment, which could be associated with arthritis, but this is also not well established. Fibrous unions and malunions are difficult to study because they are uncommon.

Proximal pole fractures: A difficult fracture type

Fractures of the scaphoid proximal to the capitolunate articulation or involving the proximal 20% of the scaphoid ( Fig. 8.5 ) are not as well studied because they are uncommon. These account for less than 5% of all scaphoid fractures. Some reports may include relatively proximal waist fractures. There is evidence that cast immobilization is associated with a 7.5 relative risk of nonunion compared to more distal fractures, but fracture displacement is not well accounted for in most studies. In one study of 52 proximal fractures, eight with displacement on CT (with the caveat that proximal and displaced were not clearly defined) diagnosed nonunion radiographically in 5 of 52 fractures, with a relatively strict CT definition of union (meaning that some of these fracture might have healed without surgery). Keeping in mind the small number of patients, there was no relationship between union and displacement.

There may be some grounds to consider early screw fixation in the hope that this will help achieve healing of the small proximal fragment with poor vascularity, but clinical data to support this are insufficient. It is true that a proximal pole fracture is more difficult to heal. However, no data have shown a surgical fixation of these fractures leads to a higher healing rate. Surgical fixations may have a nonunion rate not different from proper casting. The need of surgery for a non-displaced proximal pole fracture is an assumption thus far. Cast immobilization is a good option for undisplaced fractures of the proximal pole of the scaphoid, but it should be noted that these fractures typically take longer to heal than the scaphoid waist fractures. In casting, the cast should be secure and unmovable over forearm and wrist to firmly protect the scaphoid.