Common radiographic imaging modalities fail to accurately predict capitate morphology

There are three morphologies of the capitate based on its lunate and scaphoid articulations: flat, spherical, and V-shaped. Following a proximal row carpectomy (PRC), the capitate articulates with the lunate facet of the radius, altering contact biomechanics at the radiocarpal joint. Therefore, capitate morphology may influence contact pressures at the capitolunate articulation and influence clinical outcomes after PRC. However, it remains unclear which diagnostic imaging technique most reliably distinguishes between capitate morphologies.

We evaluated the ability of plain radiographs, two-dimensional computed tomography (2D-CT), three-dimensional (3D)-CT reconstruction, and magnetic resonance imaging (MRI) to predict capitate type in 47 fresh frozen cadaver wrists. Two attending hand surgeons and one hand surgery fellow characterized capitate type based on each imaging modality. True capitate type was determined after gross dissection. We determined the reliability of each modality to predict capitate morphology.

We found all four imaging modalities to have a low sensitivity and specificity for predicting capitate morphology. Plain radiographs, 2D-CT, 3D-CT, and MRI had sensitivities/specificities of 0.46/0.57, 0.54/0.72, 0.54/0.52, and 0.56/0.65, respectively. All modalities had high negative predictive values for detecting the more rare V-shaped capitate subtype (range 91-94 %). Inter-rater reliability was poor for all modalities.

These data suggest that plain radiographs, CT, 3D-CT, and MRI are not helpful in preoperative determination of true capitate morphology. Plain radiographs are as effective as more cost-intensive modalities in ruling out V-shaped capitates.