Hi, my name is Hayden Baker, I'll be presenting our study on sensitivity of preoperative imaging for patellar femoral instability. These are disclosures, none of them are relevant. Patellar instability is common in young athletic populations, accounting for approximately two to three percent of knee injuries. First time patellar dislocations are often treated non-operatively, up to 44% of these patients will experience recurring instability. The presence of a large chondral osteochondral defect or loose body is a relative indication for operative treatment in first time patellar dislocations. The incidence of osteochondral damage resulting in loose bodies and acute first time patellar dislocation has been reported as high as 50%. Concurrently there is no accepted standard on whether or not an MRI should be obtained in first time patellar dislocations to rule out chondral injury. There's also some concern that MRI may not identify more subtle chondral injuries or loose bodies. The purpose of this study was to define the sensitivity of MRI and x-ray in identifying osteochondral defects or loose bodies in patients undergoing operative treatment for patellar instability. Our secondary objective was to determine how often surgeons needed to address chondral pathology in patients who underwent concomitant diagnostic knee arthroscopy and prior to MPFL reconstruction. Our hypothesis was that MRI would be more sensitive and specific for detecting chondral injury compared to plain films, and that a significant portion of the patients undergoing concomitant diagnostic knee arthroscopy would have chondral pathology requiring intervention that was not identified in preoperative imaging. This was a retrospective cohort study of patients treated operatively for patellar instability over five years at an academic tertiary referral center. Inclusion criteria were evidence of clinical patellar instability with corroborating physical examination findings. Preoperative x-ray and MRI studies performed and available for review in concomitant diagnostic knee arthroscopy and MPFL reconstruction performed to address patellar instability. Patients were excluded if they had a history of prior procedure for patellar instability, underwent MPFL reconstruction without concomitant diagnostic knee arthroscopy, and had ACL or PCL deficient knees. Operative notes and arthroscopic images were reviewed to identify osteochondral injuries and loose bodies noted during diagnostic arthroscopy. Additional procedures, including chondroplasty and microfracture, were recorded. Interoperative results were then used to calculate sensitivity and specificity of preoperative imaging. Each patient's preoperative radiographs and MRI were evaluated by two independent authors for identification of intra-articular loose bodies or osteochondral defects. Disagreements were settled by the senior author. Eighty-seven patients met the eligibility criteria. Sixteen patients were excluded from sensitivity analysis because their preoperative imaging was not available for review. Final cohort included 71 patients. The majority of the patients included in the study were female, 68%. The average age at surgery was 20.6 years. For intra-articular loose bodies, we found that MRI was more sensitive than X-ray, however both were poorly sensitive. For osteochondral lesions, again, MRI was more sensitive than X-ray, but both were poorly sensitive. There was near-perfect agreement on intra-observer reliability for radiographs and identifying loose bodies, and substantial agreement for MRI. MRI had near-perfect agreement for identification of osteochondral defects compared to only substantial agreement for radiographs. Of the 87 operative reports reviewed, 45 reported performing chondroplasty for outer bridge grades two and three chondrolesions, while 23 performed microfracture for outer bridge grade four chondrolesions. Most common location for chondroplasty and microfracture was the patella. Our most significant finding was that both MRI and X-ray poorly identified loose bodies in osteochondral lesions after patellar dislocation. These findings should be considered when determining whether or not operative intervention is indicated for an acute patellar dislocation. Our findings are consistent with previously published studies demonstrating sensitivity of MRI at detecting osteochondral lesions to be 60%. The demonstrated high false negative rate of identifying loose bodies or osteochondral lesions on MRI suggests that a number of loose bodies, chondral, and osteochondral injuries are missed by MRI and X-ray. Our results suggest that treating surgeons should consider performing a diagnostic knee arthroscopy to assess for possible osteochondral defect or loose body at the time of MPFL reconstruction, as this may have been missed on preoperative imaging. The retrospective nature of this study limits our ability to derive conclusions to only previous collected data. All patients included were managed operatively, which introduces selection bias by only including patients who were experiencing severe symptoms and potentially had higher grade chondral injuries. Multiple providers at our institution perform MPFL reconstruction with varying levels of detail described in their operative reports, including loose body size, which limited our ability to perform a robust analysis of loose body size. In general, we consider a loose body to be greater than five millimeters. In conclusion, MRI and X-ray are both poorly sensitive at identifying loose bodies and osteochondral defects after patellar dislocation. For sensitivity of imaging studies must be considered when determining whether or not recommended operative management in patients with patellar instability. Thank you.

preoperative imaging

patellar femoral instability

operative treatment

MRI

x-ray