Introduction and Aims: To describe the arthroscopically assisted technique of medial patello-femoral ligament reconstruction using a hamstring tendon and evaluate the results of the procedure. Since April 2001 the senior author has performed 20 cases. One was bilateral. Average age was 30 years (range 17–52).

Method: The study group comprised nine males and 11 females. The indication for surgery was recurrent lateral dislocation in 13 and lateral mal-alignment in the remaining seven. Thirteen of the 20 had undergone prior surgery, eight an isolated lateral release and five a lateral release combined with open extra-synovial medial retinacular plication. Key features of this technique are use of the 70-degree arthroscope in the supero-lateral portal, endoscopic lateral release and use of semitendinosis or gracilis tendon attached distally. The tendon is passed through drill holes in the medial aspect of the patella and attached to a clinically isometric point near the medial femoral condyle. The tension is assessed arthroscopically and clinically before attachment. The semitendinosis tendon was used in 11 cases and gracilis in 10.

Results: All had subjective improvement with regard to patellar instability, activity level and relief of pain. There were no recurrent dislocations. The only patient in whom the tendon was attached to the distal adductor magnus required revision surgery with improvement after attachment of another tendon to bone at the isometric point. Mean follow-up was 17 months (range 4–33 months). The mean time to return to work was five weeks (range five days–12 weeks) and return to sport 10 weeks (range 3–16 weeks). The Fulkerson patello-femoral joint evaluation score improved from 46/100 pre-operatively to 74/100. Only four described the state of the knee subjectively as fair and the rest good or excellent. All patients regained full range of motion.

Conclusion: This arthroscopically assisted technique of medial patello-femoral ligament reconstruction using a hamstring tendon allows the patella to be seen and felt to articulate in the trochlea groove. It allows early rehabilitation, minimal post-operative pain, provides patella stability and significant functional improvement with rapid recovery and a low incidence of skin numbness.

Introduction and Aims: The ‘Natural Knee’ (versions I and II-Zimmer) is a primary tri-compartmental device designed for either cemented or cementless application. The cementless option includes a metal backed patella component. We present a three to seven year analysis of survival of 156 knees.

Method: Data was collected at post-operative times of one year and five years or more, to derive HSS scores. Radiological assessments included full length standing radiographs in maximum knee extension and neutral rotation, Merchant views at 30 degrees and fluoroscopically screened AP and lateral views.

Results: Of the 152 patients, 12 had died and two were lost to follow-up. One hundred and thirty-eight patients (156 knees) were available for review. Mean HSS score pre-op was 59 and at five years 91.3. Twenty-two required manipulation under anaesthesia post-operatively. Fourteen of the 122 cementless knees were revised but none of the cemented implants have required revision. Mean time to revision was 40.3 months (range 9–80 months). There were no deep infections. Revisions were for malalignment (one), secondary patella resurfacing (one), late PCL rupture (one), early tibio-femoral poly wear (two), metal backed poly fracture (two), tibial osteolysis without loosening (two), established aseptic loosening (3), and pain without loosening but hot bone scan suggesting subtle aseptic loosening (2). These last two were well fixed at revision and conversion to cemented primary components gave excellent results. There was a tendency for screw track osteolysis that was particularly difficult to detect on plain radiographs. The overall survival of this knee system at seven years when applying the Kaplan-Meier analysis is 87.4%. There was statistical significance comparing failure of cementless and cemented fixation (p=0.035).

Conclusion: This study demonstrates a significant early failure in the Cementless Natural Total Knee replacement system. Screw tract osteolyis of the tibial component is of particular concern as it is diffcult to demonstrate this on routine radiographs. Early failure of the metal-backed patella component in the absence of significant axial or rotational alignment is unacceptable.