Biomechanical consequences of tibial tubercle osteotomy in a 3D-printed model of patellofemoral dysplasia (207)

Abstract

Objectives:Trochlear dysplasia and an increased tibial tubercle-trochlear groove (TT-TG) distance are two major contributing factors to patellar instability and are often found concurrently. Patellar morphology is also abnormal in the setting of trochlear dysplasia. Indications for tibial tubercle osteotomy (TTO) include recurrent patellar instability in the setting of an increased TT-TG distance. While anteromedialization (AMZ) TTO has been shown to decrease overall PF contact stresses and improve patellar tracking, this has never been demonstrated in a model of PF dysplasia. Due in part to a lack of available dysplastic cadaveric specimens, few studies have investigated the consequences of PF dysplasia on PF biomechanics. Our previous work has demonstrated that when compared to normal morphology, PF dysplasia results in a lateral shift but negligible increases in patellar contact forces. This prompted the question of how TTO affects contact mechanics in this setting. The objective of this study was to quantify contact mechanics and kinematics following TTO using a 3D-printed PF dysplasia model. We hypothesized that an anterior tubercle position simulating AMZ TTO would best improve PF contact mechanics.Methods:Five fresh frozen cadaveric knees were dissected free of all soft tissues except the extensor mechanism. Computed tomography (CT) scan of each specimen confirmed no trochlear dysplasia or patella alta and a normal TT-TG distance (<10 mm). Dysplastic bone geometries were derived from patient CT scans selected by the senior orthopaedic surgeon who specializes in PF surgery. Segmentation was performed using Mimics (Materialise Figure 1A&B). Cadaveric knees were grouped based on the medial and lateral epicondylar distance (ML distance), and the implants were scaled to the size of each group. Scaling was done using Geomagic Studio (3D Systems), and implants were printed using a Form2 SLA 3D printer (Formlabs). Durable resin (Formlabs) was used to minimize wear between the printed components (Figure 1C). Cadaveric bony resection was performed using Biomet Vanguard (Zimmer Biomet) equipment. The amount of bone resected matched the 3D implant dimensions. A 6° distal femoral valgus cut angle was utilized. For femoral rotation, posterior referencing was utilized (no lateral insufficiency was observed), and cuts were made with 3° of external rotation in relation to the transepicondylar axis. The 3D implant was then fixed flush to the distal femur and native trochlea using screws. A metered patellar reamer was used for patellar preparation. The patellar implant was pressed into a central peg hole and fixed with a screw placed through the anterior patella. A flat tibial tubercle osteotomy cut, matching the aforementioned femoral rotation, was made with a shingle thickness of 1 cm and length of 6 cm. Each knee was mounted to a custom fixture on a servo-hydraulic load frame (MTS, Eden Prairie, MN) and cycled 5 times from 0° to 70° by pulling on the quadriceps tendon using a pulley system (Figure 1D). The shingle was fixed to the tibia using two 1.57mm K-wires. For each specimen, testing was repeated for each of three tibial tubercle positions: Native tubercle position (“normal”), 1 cm lateral to native (“lateral”), and 1 cm anterior to native (“anterior”) (Figure 2A-C). For the anterior position, a 1 cm thick plastic bone block was placed between the shingle and the tibia while maintaining its native position in the coronal plane. The lateral position was intended to represent the presurgical pathologic state (increased TT-TG), the native position a postsurgical medialized state, and the anterior position a postsurgical anteromedialized state. PF contact pressures were recorded using an electronic pressure sensor (sensor #5040, Tekscan, Boston, MA). Contact data was separated to the medial and lateral facets by identifying the median patellar ridge on the sensor. Within each facet, the sum of forces and center of pressure (weighted average of position of all acting forces within the facet relative to the median patellar ridge) was computed. Kinematics were recorded using a reflective marker motion capture system (Cortex, Motion Analysis Corporation, Santa Rosa, CA). Repeated measures ANOVA with post hoc Bonferroni analysis was used to determine differences in contact force and center of pressure location for each tubercle position. Statistical significance was defined as p<0.05.Results:There was a significant increase in the lateral facet, medial facet, and total patellar contact forces with lateral tubercle position compared to the anterior position (Figure 3). There was also a significant increase in medial facet and total patellar contact forces with the native tubercle position compared to the anterior position. There were no significant differences in lateral facet, medial facet, or total patellar contact forces when comparing the native and lateral tubercle positions. There was a trend toward an increased (lateralized) lateral facet center of pressure when comparing the lateral and anterior tubercle positions (Figure 4).Conclusions:Using a model capable of quantifying kinematics and contact mechanics for dysplastic trochleae and patellae, we demonstrated that an anterior tubercle position resulted in decreased patellar contact forces when compared to lateralized and native tubercle positions. These findings suggest that when an AMZ TTO is performed in the setting of an increased TT-TG distance and PF dysplasia, overall patellar contact forces are reduced. This may improve PF biomechanics and potentially decrease the likelihood of future PF OA. Similar findings were not observed for the native tubercle position, suggesting that anterorization is a critical consideration in improving PF biomechanics in this setting.