Return To Abstract Listing

Scientific Session 22 — SS22: Musculoskeletal Imaging - Miscellaneous

Thursday, May 9, 2019

Abstracts 2117-3049

2731. Detailed Characterization of Quadriceps Tendon Tear Patterns and Locations of Rupture: A Retrospective Study

Falkowski A1,2*,  Jacobson J1,  Kalia V1 1. University of Michigan, Ann Arbor, MI; 2. University Hospital Basel, University of Basel, Basel, Switzerland

Address correspondence to A. Falkowski (

Objective: Rupture of the quadriceps tendon is a well-known but uncommon pathology. Although quadriceps tendon ruptures have been described previously, cited studies included a low number of patients and lacked details for the specific rupture sites. Thus, our purpose is to evaluate quadriceps tendon tears with regards to which tendon components are torn, tear sites, and presence of bony involvement.

Materials and Methods: Institutional review board approval was obtained, and informed consent was waived for this retrospective study. We used the Electronic Medical Record Search Engine (EMERSE) to search our database of > 2.3 million patients for keywords “quadriceps tendon rupture,” “quadriceps tendon tear,” “knee MRI,” and their combinations. Exclusion criteria included negation of these terms, postoperative cases, and cases with no imaging available. Cases were randomized and evaluated independently by two fellowship-trained musculoskeletal radiologists. Presence or lack of a bone avulsion was documented on knee radiographs. Magnetic resonance (MR) imaging was used for measurements (length of tendon stump and tendon retraction) and to evaluate characteristics for each tendon [rectus femoris (RF), vastus medialis (VM), vastus lateralis (VL), and vastus intermedius (VI) tendon] separately: tendon quality (normal, tendinosis, partial tear, or full-width tear), location of the tear (bone avulsion, at bone insertion, proximal to the bone insertion, or myotendinous junction). Descriptive statistics were calculated. The group of patients with bony avulsions was compared to those without bony avulsions via Student’s t-tests. Interreader reliability was calculated with Cohen’s kappa and Wilcoxon signed rank test.

Results: Fifty-two patients with 53 quadriceps tears (28 right and 25 left knees) were evaluated (25 male and 7 female participants; mean age of 51 ± 13 years). The VI tendon more often incurred a partial rather than a full-width tear (39.6% versus 37.7%), whereas the RF, VM, and VL incurred full-width tears more commonly (63.2 – 66.0%). Additionally, patients with bone avulsion on radiographs had higher grade tears of the RF, VM, and VL (p = 0.020 – 0.043) but not the VI. On MR imaging, most tendons tore either at the bone insertion or proximal to the bone (89.3% of cases). In the bone avulsion group, the VM and VL tendons most often tore at the bone insertion (p = <0.001 – 0.011), whereas the other tendons showed no predilection for tear site. For each tendon, the mean length of tendon stump measured 0.7 – 1.0 cm. Torn tendon retraction ranged from 2.3 to 2.7 cm. Interreader reliability was substantial to almost perfect (K = 0.624 – 0.953).

Conclusion: The trilaminar quadriceps tendon tends to tear at its superficial and middle layers more frequently than its deep layer and usually in proximity to the patella. Scrutiny in imaging evaluation should be given to the presence of a bone avulsion on radiographs because quadriceps tendon tears tend to be more severe in this context. Knowledge of tear characteristics enables more accurate diagnosis and fosters appropriate treatment.