The Repiphysis expandable limb salvage implant system is thought to be a substantial advance in pediatric limb salvage surgery because of its innovative mechanism of noninvasive lengthening4,5. The mechanism of this implant has been described in detail previously and consists of a titanium tube embedded in a polymeric housing cylinder (Fig. 4)4,5,7. When a concentrated electromagnetic field is applied through an external coil placed around the limb, induction heating of the internal mechanism softens the polymeric tube and unlocks it from the inner titanium tube, allowing the titanium tube to longitudinally slide under the force of a precompressed cobalt-chromium (Co-Cr) spring. When the electromagnetic field is turned off, the entire system cools and the polymeric tube and the interior titanium tube are locked rigidly again in a new expanded position. Unlike other expandable implants2, this system has few internal mechanical parts, and is therefore commonly thought to be less susceptible to mechanical failure.
Although surgeons have gained considerable experience with the Repiphysis implant over the last decade, little information regarding this internal prosthesis has been published to date. In particular, there is a lack of information about the failure modes and the ability to salvage such failures. As these implants are followed for a longer period of time, failures are to be expected. Although aseptic loosening, implant fracture, dislocation of antirotation pins, and failure of the expansion mechanism have been reported4,5,8, we know of only one previous case report describing a detailed failure in a fractured expandable body of this prosthesis7. Our first two patients (Cases 1 and 2) appear to have had this mechanism of failure7.
In our opinion, the most likely sequence of events in our first two patients was the failure of the so-called lock between the trumpet flare at the end of the titanium housing tube and the polymer retaining tube. In an intact prosthesis, the flare is embedded within the wall of the surrounding polymer retaining tube. With excessive loading, slight movement between these two components can occur, resulting in progressive loss of contact between the trumpet flare and the polymer. Unless something is done to restore this lock (e.g., a lengthening that places the trumpet flare into a new region of the retaining tube), the metal trumpet flare will continue to erode the polymer tube, leading to further progressive motion. With loss of the trumpet flare-polymer cylinder lock, cyclical loading during weight-bearing allows the spring to compress and expand slightly with each step as the spring begins to carry some of the load during walking. This Co-Cr spring, which is not intended to be cyclically loaded in this fashion, may then distort and break. The distorted spring coils may also contact the thin-walled ceramic isolant ring surrounding the spring, causing it to break. If the implant is allowed to continue to deteriorate, eventually the trumpet flare will break off of the end of the spring housing tube and, along with pieces of spring, become fragmented within the distal aspect of the housing of the implant. Small metal and ceramic fragments can then move out through the sterilization vent hole in the polymer and collect around the implant and in the surrounding tissues, findings which may be visible on radiographs. This failure mode possibly had already been initiated in our first patient (Case 1), as evidenced by the extensive metallosis. Fortunately, this first patient had been scheduled for a planned revision before the clinical failure, while the second patient (Case 2) presented after implant failure.
In our third patient (Case 3), the mechanism of implant failure was different. Rather than failure of the expansion mechanism with resultant shortening, the patient had loss of the so-called restraining mechanism, leading to acute lengthening of the prosthesis. In our opinion, this was most likely due to catastrophic failure of the trumpet flare, which broke off the end of its titanium tube, allowing the precompressed spring to expand and cause acute lengthening of the implant. The very small diameter of this implant created for this three-year-old child may have contributed to the mechanical failure. Because of the smaller bone diameter and limb size, all component diameters had to be downsized during design and manufacturing. In retrospect, a single-piece, nonexpandable component with planned exchange to a so-called normal-sized expandable prosthesis at a later date may have been a better surgical choice. Wilkins and Soubeiran had earlier reported the case of a patient involving the original Phenix implant in which the polymer body sustained a posttraumatic fracture5, causing the prosthesis to disassemble with unopposed expansion of the spring, causing acute limb-lengthening. This was managed by an emergent exchange to a new Phenix prosthesis. The cause was ascertained to be a manufacturing flaw that was due to a change in the design for that particular patient.
These implant failures raise many questions about the mechanical stability and long-term durability of the Repiphysis implant. Unexpectedly, the apparent failure mode of the expansion mechanism suggests to us that unexpanded implants may be at greater risk of mechanical failure than implants that are actively undergoing expansion. This risk may increase over time, and may be due to the progressive erosion of the polymer tube by the metal trumpet, which remains in the same relative position in the tube in the absence of implant expansion. The exact time interval for this to occur is unknown. Our first two patients had their last lengthening two years and 1.5 years prior to their respective failures. Moreover, it is likely that erosion leading to this failure mechanism may be multifactorial, including the size of the implant and the forces applied to it during weight-bearing. While it is widely recognized that any type of pediatric expandable implant eventually requires conversion to an adult-sized implant, the question remains about when should such a conversion occur, particularly if a patient is functioning well with the current implant. The effect of downsizing the prosthesis in very young children, as in our third patient (Case 3), may be detrimental to the prosthesis expansion mechanism, but the question as to what should be the minimum size of this prosthesis or the minimum age and/or size of the patient for the Repiphysis usage remains unanswered. The long-term effect of extensive metallosis and exposure to wear debris in patients who have these failures is unknown. While surgeons and the patient's family need to know the potential risks of the use of this implant, our experience to date suggests that salvage of a failed or failing Repiphysis implant can be done in a relatively safe fashion4,5,7,8.
NOTE: The authors thank Robert L. Daily, BS, manager and engineer, Custom Orthopaedics, Orthopaedic Reconstruction Division of Wright Medical Technology (Arlington, Tennessee) for his expertise, assistance, and critical comments during preparation of this manuscript. The authors also thank Dhruv Kumar, MD, and D. Ashley Hill, MD, for providing the photomicrographs.