Sexual and ethnic polymorphism render prosthetic overhang and under-coverage inevitable using off-the shelf TKA implants

PurposeUnexplained pain and stiffness after total knee arthroplasty (TKA) often result from mismatch between implant geometry and knee morphology, which depends on patient size, sex, and ethnicity. This study aimed to determine whether size, sex, or ethnicity are independently associated with distal femoral morphology in healthy Caucasian and Asian knees, and to compare anatomic ratios to those of commercially available TKA implants.MethodsTwo series of computed tomography (CT) angiograms from France (264 knees) and China (259 knees) were used to digitize osteometric landmarks at the level of the femoral epicondyles, to measure anteroposterior (AP) and mediolateral (ML) dimensions at the anterior, posterior, medial, and lateral zones. The aspect (ML/AP), trapezoidicity (MLp/MLa), and asymmetry (APl/APm) ratios, as well as the sulcus angle were calculated and compared to those of 9 TKA models. Multivariable analyses were performed to determine whether anatomic ratios were independentlyassociated with sex, origin, or size.ResultsMultivariable analyses revealed that, independently from size, female knees were narrower (beta = - 0.03; p<0.001) and more asymmetric (beta=0.02; p<0.001), while Chinese knees were more trapezoidal (beta=0.04; p=0.002) and asymmetric (beta=0.02; p<0.001) with shallower trochleae (beta=6.4 degrees; p<0.001). Compared to native knees, most implants were too wide, and many of the recent models too trapezoidal'. Most prosthetic trochleae were too shallow compared to French knees, but within the third quartile of Chinese knees.ConclusionsThe morphology of the distal femur depends on sex and ethnicity independently from size. The wide spectrum of morphotypes observed cannot be covered by off-the-shelf' TKA models, and until customized implants become more accessible, prosthetic overhang, and under-coverage remain inevitable.Level of evidenceIII, Retrospective comparative study.