Diagnostic value of MRI in traumatic triangular fibrocartilage complex injuries

Key points

1. 01MRI achieved almost perfect agreement with arthroscopy for detecting any traumatic TFCC tear (kappa 0.93, sensitivity 0.99)
2. 02Localization to the distal TFCC component had only moderate agreement (kappa 0.44) and to the proximal component only fair agreement (kappa 0.21)
3. 03Peripheral tears (Palmer 1B) were the most common subtype, accounting for 67.9% of cases
4. 04Combined injuries (Palmer 1A+1B at 14%; Palmer 1B+1D at 8.3%) are frequent and not fully captured by existing MRI classifications
5. 05MR arthrography or CT may be needed as complementary tools when precise subtype classification is required for surgical planning

How it was conducted

Design

Retrospective single-center cohort study

Participants

193 consecutive patients with arthroscopically confirmed traumatic (Palmer type 1) TFCC tears, July 2020 to May 2022

MRI protocol

3 Tesla scanner (GE MR750), 8-channel wrist coil, native (non-contrast) high-resolution proton-density and T2 sequences in three orthogonal planes

Reference standard

Wrist arthroscopy performed by three hand surgeons blinded to the MRI classification; inter-rater ICC 0.95

Primary outcome

Sensitivity, specificity, PPV, NPV, accuracy, and Cohen kappa for MRI versus arthroscopy in detecting and localizing TFCC tears

Classification systems

Palmer classification (traumatic subtypes 1A-1D) and Atzei classification for peripheral tears

What they found

• Traumatic TFCC tears overall: sensitivity 0.99 (95% CI 0.97-1), specificity 0.90 (0.78-0.96), NPV 0.97 (0.87-1), kappa 0.93 (almost perfect agreement)
• Distal TFCC component (styloid lamina) tears: sensitivity 0.93 (0.88-0.96), specificity 0.53 (0.30-0.75), NPV 0.47 (0.26-0.69), kappa 0.44 (moderate agreement)
• Proximal TFCC component (foveal lamina) tears: sensitivity 0.85 (0.74-0.92), specificity 0.38 (0.29-0.49), NPV 0.79 (0.65-0.89), kappa 0.21 (fair agreement)
• Palmer 1B (peripheral) was the most prevalent subtype at 67.9% of 193 patients
• Combined Palmer 1A+1B injuries occurred in 14% and Palmer 1B+1D in 8.3% of patients
• Among 174 patients with peripheral tears, Atzei Class 1 (distal-only) accounted for 58.6% and Class 3 (complete tear) for 29.9%
• DRUJ instability was present in 37% of Palmer 1A+1B cases and 50% of Palmer 1B+1D cases
• Inter-rater ICC between the two radiologists reading MRI was 0.97

Limitations

• Retrospective single-center design limits generalizability across different institutions and countries
• Only traumatic (Palmer type 1) injuries were included; findings do not apply to degenerative TFCC lesions
• Native MRI only was used; MR arthrography and contrast-enhanced MRI may yield higher localization accuracy but were not compared directly in this cohort
• Correlation between associated injuries (e.g., lunotriquetral dissociation) and TFCC subtypes was reported descriptively but not statistically analyzed

Why it matters

For patients

A normal MRI is very reassuring for ruling out a traumatic TFCC tear, but if the MRI is positive and surgery is being planned, arthroscopy will still likely be needed to confirm exactly which part of the TFCC is torn.

For clinicians

MRI is an excellent first-line tool to confirm the presence of traumatic TFCC injury, but clinicians should not rely on conventional MRI alone to determine whether foveal reattachment or more complex repair is required; MR arthrography, CT, or arthroscopy adds necessary detail for surgical planning.

For readers

This study quantifies what many hand surgeons suspect: MRI catches the injury well but struggles with the fine-grained anatomical subtyping that drives operative decisions, reinforcing arthroscopy as the standard when localization matters.