Timely diagnosis of carpal fractures is important for optimal treatment outcome. Magnetic resonance (MR) imaging performed early in the workup can provide an accurate and specific diagnosis, as well as uncover developing complications. In the absence of carpal fractures, MR imaging can provide alternative diagnoses such as other unsuspected fractures involving the distal radius/ulna or hand, osseous contusions, and soft-tissue injuries to various ligaments about the wrist. Although MR imaging is the imaging modality of choice to assess the viability of the proximal fragment in scaphoid nonunion, the optimal imaging protocol is not universally agreed upon.
Key points
- •
The early diagnosis of carpal fractures is critical for appropriate, timely management to minimize complications related to delayed diagnosis such as fracture progression, nonunion, osteonecrosis, and secondary degenerative arthritis.
- •
In the setting of an acute wrist injury, the American College of Radiology (ACR) recommends radiography as the initial imaging modality, with a specialized carpal tunnel view in suspected hook of the hamate fractures.
- •
On obtaining negative results on initial radiographs, early magnetic resonance (MR) imaging can exclude carpal fracture.
- •
When a carpal fracture is excluded, MR imaging can also provide alternative diagnoses, such as other unsuspected fractures involving the distal radius/ulna or hand, osseous contusions, and soft-tissue injuries.
- •
MR imaging is the imaging modality of choice to assess the viability of the proximal fragment in scaphoid nonunion. Unfortunately, the imaging protocol is not universally agreed upon.
Introduction
The early diagnosis of carpal fractures is important for appropriate, timely management of these injuries. Potential complications related to delayed or missed diagnosis include fracture progression, nonunion, osteonecrosis, and secondary degenerative arthritis. There may also be a substantial cost associated with lost work productivity related to a delayed diagnosis.
The ACR has developed Appropriateness Criteria (AC) to help identify the most appropriate imaging studies that should be performed for a particular suspected condition. The AC have been developed for nearly every major organ system by a panel of experts on the subject, using all available literature to date. Based on this body of evidence, the expert panel assigned a number for the various imaging modalities indicating the appropriateness of the test in various stages of the workup of the suspected condition. A 1 to 9 rating scale is used: 1, 2, and 3 are usually not appropriate; 4, 5, and 6 may be appropriate; 7, 8, and 9 are usually appropriate. When ionizing radiation is used, the AC also provide the relative radiation levels for the imaging modalities. Fortunately, ionizing radiation imparted to the adult wrist by either radiography or computed tomography (CT) is small, on the order of 0.1 mSv. In comparison, the average yearly background radiation in the United States is approximately 30 times this amount. In the setting of an acute wrist injury, the ACR recommends radiography as the initial modality in the workup (AC rating 9). Standard 2 views are not sufficient, with up to 4 views preferred and additional views obtained as necessary depending on the suspected fracture. Unfortunately, in cases in which the carpal fracture is radiographically occult, further evaluation is required. MR imaging can provide early diagnosis of carpal fractures with the added benefit of not imparting ionizing radiation. In addition, MR imaging can provide alternative diagnoses when a carpal fracture is not identified, such as other noncarpal fractures, osseous contusions, and soft-tissue injuries, including intrinsic and extrinsic ligamentous pathology (MR Imaging of Wrist Ligaments by Ringler and colleagues). For the latter diagnoses, high spatial resolution and signal to noise ratio is required, facilitated by a field strength of 1.5 T or higher. In a recent publication, 136 consecutive patients with suspected, radiographically occult, scaphoid fractures were studied. Within 7 days of initial presentation with negative results on radiographs, 1.5-T MRI was performed. A total of 16 (11.7%) patients had scaphoid fractures, whereas 14 (10.3%) had fractures of other carpal bones or the distal radius, 55 (40.4%) had osseous contusions, and 50 (43.4%) had soft-tissue injuries.
Introduction
The early diagnosis of carpal fractures is important for appropriate, timely management of these injuries. Potential complications related to delayed or missed diagnosis include fracture progression, nonunion, osteonecrosis, and secondary degenerative arthritis. There may also be a substantial cost associated with lost work productivity related to a delayed diagnosis.
The ACR has developed Appropriateness Criteria (AC) to help identify the most appropriate imaging studies that should be performed for a particular suspected condition. The AC have been developed for nearly every major organ system by a panel of experts on the subject, using all available literature to date. Based on this body of evidence, the expert panel assigned a number for the various imaging modalities indicating the appropriateness of the test in various stages of the workup of the suspected condition. A 1 to 9 rating scale is used: 1, 2, and 3 are usually not appropriate; 4, 5, and 6 may be appropriate; 7, 8, and 9 are usually appropriate. When ionizing radiation is used, the AC also provide the relative radiation levels for the imaging modalities. Fortunately, ionizing radiation imparted to the adult wrist by either radiography or computed tomography (CT) is small, on the order of 0.1 mSv. In comparison, the average yearly background radiation in the United States is approximately 30 times this amount. In the setting of an acute wrist injury, the ACR recommends radiography as the initial modality in the workup (AC rating 9). Standard 2 views are not sufficient, with up to 4 views preferred and additional views obtained as necessary depending on the suspected fracture. Unfortunately, in cases in which the carpal fracture is radiographically occult, further evaluation is required. MR imaging can provide early diagnosis of carpal fractures with the added benefit of not imparting ionizing radiation. In addition, MR imaging can provide alternative diagnoses when a carpal fracture is not identified, such as other noncarpal fractures, osseous contusions, and soft-tissue injuries, including intrinsic and extrinsic ligamentous pathology (MR Imaging of Wrist Ligaments by Ringler and colleagues). For the latter diagnoses, high spatial resolution and signal to noise ratio is required, facilitated by a field strength of 1.5 T or higher. In a recent publication, 136 consecutive patients with suspected, radiographically occult, scaphoid fractures were studied. Within 7 days of initial presentation with negative results on radiographs, 1.5-T MRI was performed. A total of 16 (11.7%) patients had scaphoid fractures, whereas 14 (10.3%) had fractures of other carpal bones or the distal radius, 55 (40.4%) had osseous contusions, and 50 (43.4%) had soft-tissue injuries.
Imaging
At the authors’ institution, MR imaging of the wrist is commonly performed at 3 T (GE 750w, GE Healthcare, Waukesha, WI), using a birdcage coil developed in house. The MR imaging protocols used are listed in Table 1 . The full MR imaging protocol for evaluation of the wrist in suspected carpal injuries includes 3 planes (axial, sagittal, and coronal) of T1-weighted and T2-weighted with fat saturation sequences. The limited protocol, used specifically for fractures or osseous contusion of the carpal bones, generally consists of coronal and sagittal T2-weighted with fat saturation sequences of the wrist to assess for bone marrow edema signal related to injury. A T1-weighted plane, sagittal or coronal, is added to assist in assessing for a linear low-signal-intensity abnormality, connoting a fracture line. One major advantage for developing a limited imaging protocol is monetary. At the authors’ institution, the cost of a limited protocol is approximately 1 to 1.5 times that of a noncontrast extremity CT, whereas a full protocol examination is approximately 2 times the cost of a noncontrast extremity CT.
| Parameter | Axial T1 FSE | Axial T2FS FSE | Sagittal T1 FSE | Sagittal T2FS FSE | Coronal T1 FSE | Coronal T2FS FSE |
|---|---|---|---|---|---|---|
| Full protocol | X | X | X | X | X | X |
| Limited protocol | — | — | Optional | X | Optional | X |
| TR (ms) | 700–900 | 3000–6000 | 700–900 | 3000–6000 | 700–900 | 3000–6000 |
| TE (ms) | Min. | 45 | Min. | 45 | Min. | 45 |
| Echo train length | 2 | 10 | 3 | 12 | 3 | 10 |
| Bandwidth (kHz) | 32 | 32 | 32 | 32 | 32 | 32 |
| Matrix | 384/256 | 384/256 | 384/256 | 384/256 | 384/256 | 384/256 |
| FOV (cm) | 10 | 10 | 10 | 10 | 10 | 10 |
| NEX | 2 | 2 | 2 | 2 | 2 | 2 |
| Section thickness (mm) | 3 | 3 | 2 | 2 | 2 | 2 |
| Skip (mm) | 0 | 0 | 0 | 0 | 0 | 0 |
Regardless of protocol, the presence or absence of bone marrow edema signal is the key to an acute or subacute osseous injury. If there is no intramedullary T2-hyperintensity, an acute fracture and osseous contusion can be confidently excluded ( Fig. 1 A ). When a discrete linear low-signal-intensity abnormality is identified, with surrounding bone marrow edema signal, a fracture can be confidently diagnosed (see Fig. 1 B). If there is a band of bone marrow edema signal, without a discrete linear low-signal-intensity abnormality, this is consistent with microtrabecular injury or osseous contusion (see Fig. 1 C). The distinction between a nondisplaced fracture and a contusion may have consequences regarding treatment, including potential surgical intervention, length of immobilization, and follow-up. Osseous contusions do not require surgical intervention, but they may require immobilization and follow-up. One study found that, of 50 patients with osseous contusions of the scaphoid diagnosed with MR imaging, 2% progressed to a fracture despite plaster immobilization for 8 weeks. However, another study found no progression of osseous contusion of the scaphoid to fracture with 6 weeks of immobilization.
MR imaging can also help assess fracture acuity in cases in which the history is difficult to ascertain. Fractures with prominent surrounding bone marrow edema signal typically represent acute injuries. Fractures with little or no surrounding bone marrow edema are usually chronic injuries ( Fig. 2 ). Although there are no studies to date specifically on bone marrow edema and the acuity of the injury in the appendicular skeleton, there is 1 study pertaining to the axial skeleton. This group of investigators found that acute vertebral body compression fractures reliably generate bone marrow edema, whereas acute fractures that occur in distraction do not reliably generate bone marrow edema. These mechanisms should be considered in assessing for carpal fractures on MR imaging.
Scaphoid fractures
Scaphoid fractures account for 58% to 89% of all carpal fractures and can be complicated by nonunion in 10% to 12% of cases. The most common mechanism of injury is a fall onto an outstretched hand, resulting in wrist dorsiflexion, imparting palmar tensile and dorsal compressive forces on the scaphoid.
Diagnosis
The most appropriate imaging study for the initial evaluation of suspected acute scaphoid fracture is radiography, according to the ACR AC ( Fig. 3 ). Radiographs should include at least 4 views: posteroanterior (PA), lateral, semipronated oblique, and PA with ulnar deviation. In a large meta-analysis, 84% of initial radiographs correctly identified a scaphoid fracture or no fracture. About 16% of scaphoid fractures missed on the initial radiographs were diagnosed with follow-up radiography. Therefore, if the initial radiographs are negative for a suspected acute scaphoid fracture, the ACR AC recommends noncontrast MR imaging (AC rating 9), cast and repeat radiographs in 10 to 14 days (AC rating 8), or noncontrast CT without contrast (AC rating 7) (see Fig. 3 ). Noncontrast MR imaging in the evaluation of suspected, radiographically occult, scaphoid fracture has a sensitivity of 100%, specificity of 95% to 100%, and high interobserver reliability (κ= 0.8–0.96). MR imaging may also be more cost-effective, at least in the United States. Dorsay and colleagues were the first group to publish the cost-effectiveness of MR imaging in the early evaluation of radiographically occult scaphoid fractures. This group of investigators used a limited protocol 1.5-T MR imaging consisting of thin-section coronal T1-weighted and T2-weighted with fat saturation sequences at the time of initial presentation following negative results on radiographs. They found that 3 of 4 patients with clinical findings suggestive of a scaphoid fracture at clinical examination, and negative findings on initial radiographs, were potentially unnecessarily immobilized and monitored. A limited protocol MR imaging examination of the wrist at their institution cost $770, whereas the traditional practice of casting and bringing the patient back to clinic in 10 to 14 days, then removing the cast and obtaining repeat radiographs cost $677. The expense would be much greater if a diagnosis was not made at the time of follow-up, because a routine MR imaging examination was often eventually obtained if a clinical suspicion for a scaphoid fracture remained. Dorsay and colleagues also suggested that lost productivity costs for patients who are potentially unnecessarily immobilized may be great. Similar cost-effectiveness results have been shown in other non-US studies with inference of substantial productivity loss. Early MR imaging can also provide alternative diagnoses when a scaphoid fracture is not present.
