Introduction and purpose: There still remain controversies with regard to knee replacements, such as the preservation or replacement of the patellar surface. Although numerous studies have compared the clinical results of the two procedures, there have been few long-term radiological follow-up studies of the non-resurfaced patella. Our purpose is to assess this follow-up and determine the relation to clinical pain pictures of patellar origin.

Materials and methods: We carried out a clinical and radiological study of 74 PFC modular arthroplasties without patellar resurfacing with a mean follow-up of 10 years (range: 9–12 years). We measured the size and height of the patella, Wiberg classification, degree of patellar sclerosis, patellar excursion and presence and size of bone spurs. We carried out an exhaustive examination of the extensor apparatus. We related the radiological data to the clinical picture of the patella.

Results: We observed a tendency toward reduced titlting, increased height and width and diminished patellar thickness. None of these variations was statistically related to the onset of patellar pain or sustained pain.

Conclusions: Radiological variations of the non-resurfaced patella in total knee arthroplasty do not result in patellar clinical pictures.