Purpose: To study the functional outcome and survivorship of custom designed knee implants for revision and primary total knee replacement surgery where off-the-shelf prosthesis were unsuitable.

Methods: We prospectively reviewed the clinical and radiological results of 20 patients with 23 custom designed total knee prosthesis from 1991 to 2000. The indications were bone loss due to multiple revisions of total knee prosthesis and debridement for infection; peri-prosthetic fractures; bone deformity with rickets and small bones of patients with juvenile chronic arthritis. All the patients had their knee designed and manufactured in the Centre for Biomedical Engineering, University College London Medical School, Stanmore, UK. There were 4 different designs of knee prosthesis used: Condylar knee of miniature size, CAD-CAM knee, Superstabiliser and Rotating Hinges. Patients were operated upon by one senior surgeon and the Hospital for Special Surgery score taken pre-operatively, at three months, and yearly by an independent research physiotherapist.

Results: Clinical and radiological results after an average of 62.5 months (range 22 to 118 months) showed that the average Hospital for Special Surgery Score improved significantly (p=0.025) from 13.5 points (range 0–48) pre-operatively to 86.5 points (range 62–96 points). Average maximum flexion post operatively measured 86.4 degrees (range 60–122 degrees). 16 knees had excellent, 5 good and 2 poor results. Three patients had an extension lag ranging from 15–25 degrees. Only one patient with juvenile chronic arthritis needed revision at five years after the index arthroplasty.

Conclusion: The clinical and radiological results compare favorably with those who had standard knee prosthesis with similar indications. Our results support the use of a custom designed knee implant as a salvage prosthesis and an alternative to arthrodesis or amputation.

Purpose: To study the effect of retaining an abraded femoral component on the outcome of primary revision of uncemented total knee prosthesis.

Method: 66 revision procedures for failed uncemented porous coated anatomic (PCA) total knee replacement were performed in 60 patients. At review, four patients had died while two were lost to follow up, therefore 60 knees in 54 patients were included in this prospective study. The principal indications for primary revision were polyethylene wear and loosening of the tibial base plate. 14 patients had a well fixed femoral component and hence were retained while 46 patients had both the components revised. All patients were prospectively assessed prior to surgery, at three months, six months and yearly thereafter. Review comprised clinical and radiological assessment. The mean follow up was 8.4 yrs (7–12 yrs)

Summary of results: The mean pre-operative Hospital for Special Surgery knee score after the first revision improved at a follow up of 8.4yr (7–12yr) giving 39 (65%) excellent, 17 (28.4%) good and four (6.6%) poor results. 13 out of 60 knees needed a further revision. Six of these second revisions which had only the tibial component changed failed very early (mean of 2.1 yr) when compared to the rest (6.8yr), probably due to wear between the abraded retained femoral component and polyethylene insert. The estimated odds ratio (relative risk) of second revision in patients with retained femoral component compared to those with revision of both the components was 4.17 (95% C.I. 1.07 – 15.4).

Conclusion: We recommend exchange of all the components at the time of revision of PCA knee prostheses. Significance: To our knowledge, this is the only study involving such a large number of PCA revisions and looking into the effect of retaining the femoral component compared to changing both the components on the outcome of revision.