This retrospective single-center study was approved by the institutional review board.
Patient selection
All patients, at least 18 years old, with radiographs and MR images of one finger acquired between January 2014 and November 2018, were selected through a search of our radiology information system database. The resulting 133 patients were reviewed manually for radiographs and MR examinations of the thumb and for injury of the UCL by reading the radiology reports. Figure 2 demonstrates the selection of patients. The medical records of the remaining 28 patients were investigated for a trauma setting and the time interval between imaging by one musculoskeletal fellow radiologist.
Patients with evidence of sprain or rupture of the UCL on MR images, a time interval of up to 6 weeks between radiographs and MRI, and imaging within 2 months of trauma were included in this study. Three patients were excluded due to the following reasons: two for having a time interval of more than 2 months between imaging and trauma, and one for insufficient image quality on MRI. Twenty-five patients met the inclusion criteria.
Imaging technique
Radiographs included a lateral and anterior-posterior view of the thumb.
MRI of the thumb was performed at 3 T (Prisma, Verio, Skyrafit, or Skyra, Siemens Healthcare) with a dedicated 16-channel wrist coil (turbo spin echo sequences, FOV, 70–133; matrix, 256–384 × 269–488). The MRI protocol included coronal intermediate-weighted (IW) fat-saturated (FS) (TR/TE, 2580–4800/42–51 ms; slice thickness [ST], 2 mm; echo train length [ETL], 9), sagittal IW FS (TR/TE, 2400–3750/37–67 ms; ST, 2 mm; ETL, 8), axial T2-weighted (TR/TE, 3800–5630/63–83 ms; ST, 2 mm; ETL, 14), and axial T1-weighted (TR/TE, 473–1000/13–14 ms; ST, 2 mm; ETL, 2) images. Additionally, one patient had a computed tomography of the thumb.
Imaging analysis
All radiographs and MRI studies were retrospectively reviewed by one fellowship-trained musculoskeletal radiologist with more than 7 years of experience (A. H.; Reader 1 [R1]) and one musculoskeletal fellow radiologist (S. M.; Reader 2 [R2]). The two modalities were interpreted independently and randomly with a time interval of 4 weeks.
Prior to the study read-out, four radiographs and MRI were analyzed by both radiologists in consensus to familiarize with the appropriate criteria. These training exams were not part of the present study.
Radiographs were evaluated for the presence or absence of avulsion fractures about the first MCP joint at the following locations: UCL, radial collateral ligament (RCL), checkrein ligaments at the neck of the first metacarpal bone, and phalangoglenoid ligaments/volar plate at the base of the proximal phalanx (Fig. 2).
The distance of dislocated fragments of collateral ligament avulsions from the donor site was measured (mm) on the anteroposterior view in PACS.
MR images were assessed for the presence or absence of a sprain, a partial or complete tear, or an osseous avulsion of the UCL and RCL of the first MCP joint.
A sprain was defined as slight hyperintense signal in IW FS images of and surrounding the ligament without interruption of fibers, a partial tear was defined as an incomplete continuity of the ligament, and a complete tear was defined as interruption of all fibers.
Dislocation was measured from the distal stump to the donor site (mm) for the UCL and RCL.
The presence or absence of a Stener lesion of the collateral ligaments was recorded. A Stener lesion was defined as a displaced ligament proximal and superficial to the adductor aponeurosis.
The accessory UCL and RCL and the dorsal ligamentous complex (capsule, extensor hood, and extensor tendon) were assessed for the presence or absence of a sprain or tear and similar criteria were applied as above. A tear of the transverse intersesamoid fibers was assessed on axial images and confirmed on sagittal images.
Similar to the radiographs, the checkrein ligaments at the neck of the first metacarpal bone and the phalangoglenoid ligaments/volar plate at the base of the proximal phalanx were assessed for the presence or absence of a sprain or tear.
Bone marrow edema pattern about the articulating structures of the first MCP joint was also documented.
Clinical and intraoperative report review
Data from the electronic medical record were retrieved for trauma history, date of trauma, final clinical diagnosis, and final treatment by the radiology fellow (S.M.).
Statistical analysis
Descriptive statistics were used to report the imaging interpretation.
A McNemar test was performed to determine statistical difference in frequencies of volar lesions between patients with a dislocated UCL tear and non-displaced UCL tear. A p value of < 0.05 was considered statistically significant.
Interrater agreement was assessed using kappa (κ) statistics for qualitative data and the intraclass correlation coefficient (ICC) for quantitative data. According to Landis and Koch, a κ value of 0–0.20 indicates slight agreement; 0.21–0.40 fair agreement; 0.41–0.60 moderate agreement; 0.61–0.80 substantial agreement; and 0.81–1 almost perfect agreement [17]. The quality of interrater reliability by means of ICC was classified as follows: > 0.75 excellent, 0.4–0.75 fair to good, and < 0.4 poor [9]. For all analyses, statistical software (SPSS version 22, IBM) was used.