A Comparative Study of Thumb Metacarpal Morphology and Shape Categorization

The authors obtained approval from the local institutional review board (IRB# IRB00004115) for this study. The current procedural terminology codes 26541 and 26540 and the International Classification of Disease codes 842.12, S63.641A, and S63.642A were used to identify patients with injuries to their thumbs (both UCL and non-UCL injuries) who were seen at Vail-Summit Orthopaedics and Neurosurgery between 2006 and 2020. Medical record numbers and the Vail-Summit Orthopaedics and Neurosurgery billing database were used to access patient radiographic images. A lateral radiograph was identified by superimposition of the condyles of the distal ends of the proximal phalange. Patients were included in the study if they met the age criteria (range, 18–75 years) and had appropriate radiographic images of the thumb MC. Patients with severe MCP joint arthritis or a lack of a clear proximal phalanx-MC condyle alignment in the lateral radiograph were excluded from the study. Thumb radiographs were uploaded using MedStrat software (Medstrat, Inc, Downers Grove, IL). The measurement and classification of shape was performed by 2 independent investigators (KRB and ED), including a board-certified hand surgeon, according to previously described criteria. Metacarpal head shapes were classified as either flat or round by dividing the distance measured from the volar to the dorsal edge of the articular surface (anteroposterior distance) by the radius of curvature of the articular surface. The radius of curvature was measured with the best fitting circle that encompassed the entire MC head and had lines parallel to the dorsal and volar side of the articular surface. Metacarpal heads with a ratio greater than or equal to 1.7 were categorized as round, and MC heads with a ratio of less than 1.7 were categorized as flat. An inter-rater reliability analysis using Pearson correlation was conducted for the 2 independent investigators (KRB and ED), which revealed high-inter-rater reliability for shape classification (r = 0.965). During the measurement and classification process, it was noted that a subset of MCs were not able to be categorized as either flat or round using the measurement method, due to an inability to appropriately determine the radius of curvature of the MCP joint on the lateral radiograph. This subset of MCs had a biplanar shape, identified by the presence of gaps between the best fitting circle and the dorsal or volar side of the articular surface, and so, these were categorized as biplanar.

A total of 318 patient radiographs were obtained using the collection methods described above. Of these 318 participants, 210 had a proper lateral radiograph, met the age criteria, did not have severe MCP arthritis, and were included for further investigation. Among the 210 study participants that were included, 110 (52.4%) were women, and the average age was 47.3 years (SD, 15.0 years). Of these participants, 113 (54%), 79 (37%), and 18 (9%) were classified as having round, flat, and biplanar MC head shapes, respectively (Table 1).

Previous studies have described MC heads as having either a flat or round morphology.^, During our primary investigation into the relationship between MC head shape and the prevalence of UCL injuries, we have identified a potential third, biplanar MC head shape. The biplanar head shape is triangular and has 2 distinct planes on the articular surface. These 2 planes occur on the dorsal and volar sides of the MC head and converge into an apex.

These biplanar MC heads were identified due to an inability to measure the radius of curvature according to previously described criteria. For MC heads to be appropriately measured, the radius of curvature of the best fitting circle must be parallel to the entire articular surface (Fig. 2A, B ). When applying this method to the biplanar MC heads, the triangular appearance prohibits the radius of curvature of the best fitting circle from being parallel to both intersecting planes. When we apply the circle to encompass the entire MC head, it results in gaps between the best fitting circle and the dorsal or volar side of the articular surface (Fig. 2C). The presence of these gaps between the best fitting circle and articular surface is what led to the visual identification of the biplanar morphology.

The biplanar MC head morphology has likely been observed before but, to our knowledge, has not been categorized in the manner that we have attempted in this study. Joseph described 2 intermediate MC head shapes as being “roundish or flattish.” When reviewing the photographs provided from that study, it appears that 1 of the radiographs of a “roundish” MC head morphology resembles what we have categorized in our study as a biplanar MC head shape. Although 90% of the MC head shapes in the Joseph study were classified as “round” or “roundish,” the authors did not identify the exact number of MC heads in each category. Our findings suggest that there may be a third distinct, biplanar MC head morphology.

We do not believe our observation of a biplanar metacarpal head to be a result of variation in radiographic projection. In a radiographic and cadaveric study, Joseph used lateral radiographs from 16 thumb MCs to understand how the radiographic projection influenced the appearance of the MC head shape. They discovered that MC heads categorized as round appeared round upon visualization in all 5 radiographic orientations and that flat MC heads could appear round on 1 or more views. Considering this, we would not expect that a round or flat MC head would appear to be biplanar on any view.

Many limitations exist, given that the study findings were secondary to our primary objective of understanding the relationship between the MC head shape and the prevalence of UCL injuries. The first and most important limitation is related to the data collection methodology, which was designed to address our primary research question and not that of this study. The lateral radiographs required to determine the shape of the MC head were only taken after a patient had sustained a thumb injury, therefore introducing a risk of selection bias into our sample. Our data collection methods also yielded a small, underpowered sample size that prohibited us from making any statistical comparisons of patient demographics and characteristics between different groups. Future studies should be designed and conducted to capture a more diverse and larger sample so that our observations regarding MC shape can be validated. More research is required surrounding MC head shapes and categorizations before any meaningful clinical application may be considered.

Funding for this work was provided by the Vail-Summit Orthopaedic Foundation. We thank the donors of the Vail-Summit Orthopaedic Foundation.