Revision of broken knee megaprostheses: new solution to old problems

Abstract

Background: Low-cost indigenous megaprostheses used in the developing world are prone to mechanical failure but the frequency and causes are not well established.

Questions/purposes: We retrospectively analyzed the causes of failure, particularly design, and suggest changes to reduce the breakage. We also report our experience with revision surgery.

Methods: We identified 28 breakages in 266 megaprosthetic knee arthroplasties performed between January 2000 and December 2006. Twenty-six breakages were revised to another prosthesis. The complications were studied and the function was evaluated. Prostheses were studied for failure by the computer-aided design program SolidWorks(®) and Hyperworks(®) for finite element analysis (FEA). Design improvements were performed based on these results.

Results: In 21 cases, the failure occurred at the stem-collar junction, the point of maximum stress predicted by FEA. Stainless steel implants were prone to failure. There was one early and one late infection. Three patients died of metastatic disease. The most difficult surgical step involved the removal of the well-cemented broken stem from the intramedullary canal. Musculoskeletal Tumor Society scores varied from 27 to 29 after revision. FEA revealed stress could be reduced by filleting the stem-collar junction and by two-piece stems.

Conclusions: Revisions of broken total knee megaprostheses, though technically difficult, have allowed patients reasonable function. We recommend design analysis for custom prostheses to point to areas of weakness. Breakages can be reduced by using titanium stems and filleting the junction or by having two-piece inserted stems. Incorporating these changes has reduced the failures in our experience.

Fig. 1A–E

Photographs illustrate the evolution of the indigenous prosthesis. (A) In the year 2000, a simple stainless steel fixed-hinge customized implant was made to start our limb salvage program. Note the sharp edges and the “squarish” condyle. The dimensions of the condyle portion and tibial tray were standardized and kept constant, with only the resection length and intramedullary stem diameter variable. The condyle size was kept small to fit all ages of patients and to facilitate easy closure. (B) In 2001, the condyle was made more rounded. Note the transverse grooves on the intramedullary stem. The grooves were the site of breakage and were eliminated in (C) 2002. The intramedullary stem was now smooth and curved to match the anatomy. This resulted in very early loosening and we reverted to matt-finished stems with a longitudinal groove in (D) 2003. Bushings and bumpers were added to prevent metallosis. The central locking hinge pin failed in a few cases and it was a problem to access the notch in revisions due to the tight quadriceps. (E) The locking mechanism was changed to a circlip in 2005.

Fig. 2A–D

Various stem designs modeled using SolidWorks^® software are shown: (A) stem with sharp transition from body; (B) stem with a fillet; (C) stem with a chamfer; and (D) a two-piece stem connected with a taper. The body is 25 mm and stem is 10 mm.

Fig. 3

Actual physical testing was performed on stems fabricated in stainless steel 316L using the geometry and shapes shown in Figure 2. The strain gauges on the stems measure the strain.

Fig. 4A–G

Photographs illustrate the steps in extraction of a broken stem. (A) A ring osteotomy is performed about 2 cm above the bone end to expose (B) the embedded stem. (C) The stem is now exposed. (D) A high-speed burr is used to remove the cement at the mouth around the stem. (E) The stem is now held with a vise grip and hammered out. (F) If unsuccessful, then a corticotomy is performed, and if the stem is still stuck, a window is made to clear the cement. (G) A postoperative radiograph shows the window put back and held with cerclage wires.

Fig. 5A–E

Sites of stem failure are shown. (A) Breakage through the transverse groove was seen in two cases. This groove was eliminated after 2001. (B) The stem-collar junction is the highest stress area, which failed in 22 cases. This figure shows the failure at the base of the fillet. (C) Failure through the body is illustrated here. The thick threaded stem has transferred load to the thin wall of the body, which has failed. This is a manufacturing flaw that has now been eliminated. (D) Failure through the tibial stem is seen in this rare instance. (E) Failure through the hinge wall in a proximal tibia is seen here. This happened in a patient who was revised 4 years earlier for failure at the stem-collar junction.

Fig. 6

A graph shows physical strain variation against applied loads. Fillet and chamfer designs show resistance to deformation than stems with sharp transition. Though the Two-piece stem shows more strain, it is distributed over a wider area improving the safety.

Fig. 7

The modular prosthesis used from year 2006. The condyles and tibial base plate are from chrome-cobalt alloy and the intramedullary stems are made from titanium.