Preoperative computed tomography scanning and arthroscopy in predicting outcome after lateral retinacular release

We initiated a study to look at preoperative, flexed-knee, midpatellar computed tomography (CT) scans and intraoperative arthroscopic findings of lateral patellar articular degeneration in predicting the results after lateral retinacular release for failed nonoperative treatment of anterior knee pain. Twenty patients with 30 painful knees underwent preoperative flexed-knee, midpatellar CT scans that were retrospectively classified by the method of Fulkerson into normal alignment, lateral subluxation, lateral patellar tilt, and combined tilt and subluxation. Arthroscopy was performed before open lateral release. The lateral facet of the patella was graded as either minimal changes (Outerbridge I or II) or advanced (Outerbridge III or IV) changes. Patients were followed for a minimum of 2 years and graded on a standard patellofemoral rating scale. Only 22 of 30 knees that were thought to be clinically malaligned, actually were malaligned by CT scan; eight CT scans were interpreted as normal. The results were further stratified into group A (CT-documented tilt, minimal facet changes), group B (CT-documented tilt, advanced facet changes), and group C (normal CT). Ninety-two percent of group A were rated good or excellent. Twenty-two percent of Group B rated good/excellent, 33% fair, 44% poor. Only 13% of group C rated good (one patient). Based on the results of the study, we recommend lateral release for anterior knee patients with CT-proven patellar tilt who have not responded to conservative treatment and have minimal facet changes with minimal or no subluxation. Lateral retinacular release should not be offered as a treatment to the patient with a normally aligned patella because poor results will most likely result.