Hypothesis: 1. Increased wear results from modularity. 2. Modification of the patello-femoral articulation will decrease patellar fractures.

Materials & Methods: A prospective comparison of 100 consecutive Non-modular Insall Burstein Posterior Stabilized (IBPS) knee prostheses (1986–1989) with 100 consecutive modular IBPS II knee replacements (1989–1990). No patient was lost.

Results: IBI: Nine re-operations of which 6 were complete revisions: two for sepsis, one for tibial loosening, one for patellar wear and two for undiagnosed pain. Of seven patellar fractures, five required surgery. IBII: Nineteen re-operations of which six were complete revisions: two for sepsis, one for tibial loosening from catastrophic osteolysis, two for instability and one for stiffness. There were no revisions for patellar complications. Of 12 non-revision re-operations, two were for patellar fractures, three for dislocation of the posterior stabilized mechanism and one for failure of the modular locking mechanism. Six knees suffered patellar “clunks” treated arthroscopically.

Discussion: The femoral component patellar groove was smoothed and the posterior stabilized mechanism was relocated on the IB II. This increased motion and decreased patellar fractures but caused scar on the quadriceps tendon to “clunk” and lead to dislocations of the “spine and cam”. All failures occurred in the first five years.

Clinical Relevance: Specific patellar problems were improved with design modifications. New problems have been addressed and long-term survival has not been compromised by modularity

Introduction: The modular IBPSII prosthesis was introduced in 1989 with modifications to the patello-femoral articulation and the posterior stabilized mechanism.

Methods: 100 consecutive IBPSII knee arthroplasties were followed prospectively. Age, gender, deformity and diagnoses were comparable to previous groups.

Results: Fifty-one knees were evaluated at 10 or more years with the Knee Society scores and radiographs. 14 were evaluated by phone. An additional 6 knees required revision and 29 were in patients who died. None were lost. Revisions were performed for instability (2 knees), sepsis (2), loosening from osteolysis (1), and stiffness (1). In the 10-year group, 12 patients required reoperations: Patellar revision for loosening (1), patel-lectomy for fracture (1), polyethylene exchange for dislocation of the spine and cam mechanism (3) and for dissociation (1), and arthroscopic resection of scar from the quadriceps tendon (patellar clunk) in 6 knees.

Conclusion: The smoother patello-femoral groove was associated with fewer patellar fractures, but resulted in scar on the quadriceps catching on the femoral component. The tibial spine was moved posteriorly from previous models to increase rollback. This resulted in dislocation of the spine and cam mechanism. One case failed due to loosening and extensive osteolysis presumably associated with modularity. The last two complications were not observed with earlier versions of this prosthesis. All complications occurred within the first five years.