Abstract
Single-piece silicone implants dominate metacarpophalangeal joint arthroplasty. The NeuFlex® implant was introduced to improve on the clinical performance of other silicone implants by having a pre-flexed hinge. By visually examining a cohort of 30 explanted NeuFlex® metacarpophalangeal joint prostheses we sought to identify the failure modes of these implants. Seven were not fractured, 11 had fractured across the hinge, nine had fractured at the junction of the distal stem and the hinge, and three showed fractures at both the hinge and at the junction of the distal stem and the hinge. These data may prove helpful in identifying how the performance of single-piece silicone implant designs can be improved.
Introduction
Single-piece silicone implants dominate arthroplasty of the metacarpophalangeal (MP) joint. The Swanson (Wright Medical, Memphis, USA) implant is the market leader (Norwegian Arthroplasty Register, 2016). Known ‘problems’ include fracture (Kimani et al., 2009), primarily at the junction of the distal stem and hinge (Joyce and Unsworth, 2002). This position of fracture is also seen with the Avanta/Sutter (Sutter Corporation/Avanta, San Diego, CA, USA) single-piece silicone finger implant (Joyce et al., 2003). It is thought to be particularly related to the dominance of subluxing forces in the rheumatoid MP joint (Drayton et al., 2016; Joyce, 2009). In 1998, the pre-flexed NeuFlex® (DePuy Synthes, Warsaw, IN, USA) implant was introduced. The three designs, Swanson, Sutter and NeuFlex®, are shown in Figure 1.
Different designs of silicone single-piece finger implants. Top, NeuFlex®; middle, Sutter/Avanta; bottom, Swanson. The distal stem is to the right in all examples.
It is recognized that fracture of silicone MP implants is common. Trail et al. (2004) reported that two-thirds of 1336 Swanson implants had fractured at 17 years follow-up. Goldfarb and Stern (2003) reported that 67% of their Swanson implants and 52% of their Sutter implants had fractured at 14 years follow-up. A fractured implant does not necessarily mean a clinical failure, but recurrent symptoms are more likely with fractured implants (Trail et al., 2004).
Clinically, it has been reported that the NeuFlex® MP joint has produced an increased range of motion compared with use of the Swanson implant (Delaney et al., 2005). This finding was also reported by Escott et al. (2010), although their patients treated with a Swanson implant had better self-reported function and aesthetics. Namdari and Weiss (2009) reported 1/29 fractures in their cohort of NeuFlex® implants. Similar fracture rates between the NeuFlex® and Swanson designs have been reported (Kimani et al., 2009). When NeuFlex® and Sutter implants were compared in the MP joint, two out of 78 NeuFlex® and five out of 78 Sutter implants broke (Pettersson et al., 2006).
The aim of this study was to assess the failure of explanted NeuFlex® implants and compare the results with established patterns for other one-piece silicone implants.
Methods
NeuFlex® MP joint explants were gathered as part of an ongoing retrieval programme. The implants were part of a consecutive series of explants for which patient demographic data were available. All revision procedures were undertaken for clinical problems, that is symptomatic joint replacement failure requiring revision surgery after failure of non-operative treatment. After removal, the implants were cleaned with a brief immersion in chlorhexidine for 30 mins and then washed in tap water. The explants were photographed using a digital camera. The position of any fracture was noted. We recorded patient demographics including the age, gender, underlying diagnosis, site of surgery and time in situ.
Based on the explant results (see below), we compared the thickness of the central hinge section of NeuFlex® and Sutter/Avanta implants since a thinner section would probably fail more quickly, as crack growth resistance is probably key to the longevity of these single-piece, silicone prostheses (ASTM-F1781-15, 2015; Hutchinson et al., 1997). Size 20 and size 40 implants of the NeuFlex® and Sutter/Avanta were measured using a Vernier caliper (Mitutoyo, Huddersfield, UK).
Results
Data from 30 NeuFlex® MP joint explants.
RA: rheumatoid arthritis; OA: osteoarthritis.
Seven explants were not fractured (Figure 2(a)). Eleven explants had completely fractured at the hinge (Figure 2(b)). A further nine explants had completely fractured at the junction of the distal stem and hinge (Figure 2(c)). Three explants had completely fractured at both the hinge and at the distal stem (Figure 2(d)). We saw a mix of fractured and intact implants in some patients, for example explants 8–11 came from the hand of one patient. One explant was intact, two had fractured at the distal stem and one had fractured at the hinge. Separate to this we saw an explant (No. 19) with an incomplete fracture.
(a) An un-fractured NeuFlex® explant (explant No. 15). (b) A NeuFlex® explant (explant No. 28) showing fracture at the hinge. (c) A NeuFlex® explant (explant No. 24) showing fracture at the junction of the distal stem and hinge. (d) A NeuFlex® explant (explant No. 13) showing fracture at both the hinge and at the junction of the hinge and the distal stem.
NeuFlex® MP explants ranged in size from 0 to 40. As might be expected, smaller sizes were retrieved from the smaller fingers of hands, whereas larger implants tended to come from the index and middle fingers. The age at revision ranged from 43 to 81 years (median 58). The time in situ ranged from 6 to 120 months (median 59). In all but two cases the diagnosis was rheumatoid arthritis. Discolouration of some explants had occurred (Figure 3). The significance of this is unclear, but it could represent some change in material properties of the implant.
Two NeuFlex® explants (explant No. 18 above and explant No. 13 below) showing discolouration. Note that discolouration of explant No. 13 has occurred throughout the entire material and not just on the outer surfaces.
The size 20 NeuFlex® and Sutter/Avanta implants were each measured to have a hinge thickness of 1.9 mm and the size 40 implants were each measured to have a hinge thickness of 2.3 mm.
Discussion
As shown in Table 1, 11 NeuFlex® explants had fractured across the hinge (Figure 2(b)); this has not previously been reported in vivo. Fracture at the hinge occurred when three NeuFlex® implants were tested in vitro in a finger function simulator (Joyce and Unsworth, 2005). Nine NeuFlex® explants had fractured at the junction of the distal stem and hinge (Figure 2(c)). This is similar to the site of implant fracture seen with Swanson implant fractures (Gellman et al., 1997; Trail et al., 2004) and Sutter/Avanta (Joyce et al., 2003) implant fractures. In vitro tests have shown that Swanson (Joyce and Unsworth, 2000) and Sutter (Joyce et al., 2003) implants fracture at the junction of the distal hinge and stem, in both cases matching clinical experience.
Intriguingly, 10% of the cohort fractured at the hinge and at the junction of the distal stem and hinge (Figure 2(d)). The authors are not aware of such a failure mode having been reported previously. That it happened in three explants, and that one of the un-fractured NeuFlex® explants (No. 19) also showed substantial damage at both the hinge and the junction of the distal stem and hinge, indicates that the failure mode was not an anomaly; of greater note 11/30 fractured just at the hinge. Overall 14/30 suffered fractures of the implant hinge. Fractures of silicone implant hinges have not previously been reported. It is unclear whether this occurs before, after or separate from stem fractures.
The one incomplete fracture (No. 19) showed that fracture began dorsally. This matches with the fracture initiation site of silicone MP joint implants seen previously (Joyce, 2009). Discolouration of most explants occurred (Figure 3) probably indicating that the silicone material can change when in the body. This may affect the material properties of the implants and could therefore be a useful area for future research. We appreciate that fractures would probably have occurred sometime before removal of the implants and therefore the time in situ does not equate with time to fracture.
Ex vivo analysis of finger implants potentially highlights key areas of failure, thus providing information that could be used to reduce failures by improving future designs. For example, based on the empirical evidence provided in this article and by previous publications (Drayton et al., 2016; Joyce, 2009; Joyce et al., 2003), subluxing forces need to be minimized, to reduce the shear stresses on the implant at both the junction of the distal stem and hinge and probably at the hinge. In addition, consideration could be given to increasing the thickness of the hinge to reduce the time taken to fracture, although there may be a concomitant increase in stiffness of the implant.
With improved designs, patients could achieve more with their replaced joints while increased longevity of prostheses should lead to fewer revision operations. The latest Norwegian Arthroplasty Register shows that revision MP joint arthroplasties accounted for 42% of all MP joint replacement operations in 2015 (Norwegian Arthroplasty Register, 2016). Therefore, revision MP joint arthroplasty is common and opportunities to reduce such operations are substantial.
There are limitations to this research. There is no denominator to show the percentage of implant failure. The details of the original operations are unknown; there may have been specific factors leading to implant failure. There is no correlation with the biomechanics of the failed joints.
This study provides new data on the failure of NeuFlex® implants, raising questions about their design and its possible link to mechanical failure.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.