Summary

Diagnosis of patella alta is difficult to determine, as dynamic and static imaging modalities result in significant differences when measuring patella-trochlea overlap

Abstract

Background

Patella alta is defined as abnormally high riding patella in relation to the femoral trochlea. Patella alta has been described as one of the most common risk factors for patellar dislocation. To stabilize the dislocating patella, patella alta may need to be addressed in the surgical management strategy. Classic indices that define patella alta, such as Insall-Salvati (IS) and Caton-Deschamps (CD), are based on native radiographs. MRI has been used for a more precise analysis of the overlap between patellar and trochlear cartilage (Patella-Trochlea Index, PTI) to define patellar engagement to the trochlea, which can be considered as a definite patella alta assessment. The reliability of MRI patella alta measurements needs to be better understood. This study aimed to compare static MRI patella alta measurements to dynamic Computed Tomograph (CT) measurements to assess patellar height. The study hypothesis was that static patella alta measurements may vary from dynamic assessment of patellar height.

Methods

From patellar dislocation population, 29 patients underwent static 3.0T MRI (supine, standard knee coil) and dynamic Conical-Beam CT (CBCT) in one leg standing position, zero degree extension, and quadriceps muscles contracted (patellar tendon tightened). Open platform CBCT allows the patient to stand on the indicated limb as in a normal stance, and imaging is finished within a few seconds. Resolution of CBCT allows analysis of patellar and trochlear cartilage overlap similar to MRI. Patellar height in relation to femoral trochlea was measured as overlapping of the patellar cartilage and femoral trochlea, described as PTI, < 15% indicating abnormal PTI and patella alta. CD and IS indices were measured as originally described. Additionally, the Lateral Trochlear Inclination (LTI) angle was measured to assess concomitant trochlea dysplasia and whether it affects to patellar height between MRI and CBCT. Crude comparison between MRI and CBCT was performed with paired sample t-test and adjusted comparison in relation to LTI with linear mixed model analysis.

Results

Patellar height measurements resulted in lower values in standard static MRI compared to dynamic CBCT, when PTI or CD was used. Most importantly, PTI was mean of 28% lower (confidence intervals [95% CI] 21-35%, p<0.01) in dynamic CBCT than static MRI, indicating significantly more frequent patella alta diagnosis in dynamic assessment than measured on static MRI. CD was mean 10% less (95% CI 7-14%, p<0.01) in MRI than dynamic CBCT, indicating similarly more frequent patella alta diagnosis when measured by dynamic imaging modality. IS was similar in both imaging modalities, indicating an invaluable method for patellar alta assessment in terms of patellar engagement to the trochlea. The distal cartilage margin of the patella in sagittal view was mean 6mm (range 2.4-12.7mm) proximal in dynamic imaging than static. The presence of trochlear dysplasia (decreased LTI <11) did not affect the measurement (PTI, CD and IS) behavior between static and dynamic imaging, indicating a similarly increasing amount of patella alta diagnosis when assessed by dynamic CBCT.

Conclusions

Diagnosis of patella alta is difficult to determine, as dynamic and static imaging modalities result in significant differences when measuring patella-trochlea overlap. Standard static MRI assessment of PTI results in significantly higher values than dynamic CBCT, indicating more frequent patella alta diagnosis when dynamic imaging modality is used. Dynamic imaging modality may be considered more precise method to assess patellar engagement within the trochlea. Care should be taken to diagnose patella alta solely based on MRI.