Editorial

The thumb carpometacarpal (CMC) joint is the most common site of osteoarthritis in the hand. Trapezectomy is an established surgical treatment (Salem and Davis, 2012). Although most patients achieve good pain relief, their pinch strength typically remains weak. Various soft tissue procedures have been performed to improve the outcome of trapezectomy. Although well established, none has reliably been shown to give a better outcome than trapezectomy alone (Salem and Davis, 2012; Vermeulen et al., 2011). Implant arthroplasty has been tried as an alternative. Various materials have been used, including silastic and hard-bearing materials, and various designs, including spacers, hemi-arthroplasties, and total arthroplasties (Vermeulen et al., 2011). There are very few long-term results (Johnston et al., 2012; Vermeulen et al., 2011) and, for some implants, very limited (van Rijn and Gosens, 2012) or no obvious published data of good outcomes (Adams et al., 2009).

In this issue we report on a number of thumb arthroplasties. We report good long-term outcomes for the de la Caffinière implant, but poor outcomes for the Moje, Elektra, and Pi2 thumb CMC joint implants. The Moje implant was first available in the mid-1990s. There appear to be no published papers reporting good outcomes with the use of the Moje thumb CMC joint implant, and the company has supplied none despite a request. Previously, Hanson and Vainorius (2008) reported a failure rate of eight out of nine implants at 1 year, primarily due to failure of implant fixation to bone. In this issue, Kazsap et al. (2012) also report a similarly high rate of failure, specifically with 75% of implants requiring or being recommended revision surgery.

For the Elektra implant there is one published paper reporting relatively good outcomes (in 83%) from the originator of the implant (Regnard, 2006). Nonetheless, in the 100 cases reported at a mean follow-up of 54 months there were seven per-operative fractures, seven post-operative dislocations, two cases of early subsidence of the metacarpal implant, and 15 cases of loosening of the trapezial implant. There appear to be no other published papers reporting good results, and the distributors have been unable to supply any other studies of good results. The only subsequent reports, published in this journal, report very high failure rates. In February 2012, Hernández-Cortés et al. (2012) reported radiological evidence of loosening in 47% at 2 years follow-up and in this issue Klahn et al. (2012) report a revision rate of 44% at 6 years.

The Pi2 (Pyrocarbon) spacer has been available recently. There have been no reports of the outcome of implanting P2 spacers until now (Maru et al., 2012). This study reports a high complication rate with six of 18 patients needing multiple operations typically for dislocation, yet with no obvious benefit over simple trapezectomy.

Metal-on-metal hip replacements have been shown to have unacceptably high failure rates (Smith et al., 2012) leading to withdrawal of implants and potentially extensive costs to the companies to fund revision operations, yet the failure rates are only around 6% at 5 years — a much lower rate than reported for many thumb CMC joint arthroplasties. Following failed Moje CMC joint arthroplasties, excision of the implant gives results comparable to a primary trapezectomy (Kazsap et al., 2012). Nonetheless, the reported failure rates of the CMC joint arthroplasties are vastly greater than the “unacceptable” failure rates of some metal-on-metal hip replacements, and yet they continue to be implanted. For the Moje CMC joint arthroplasty, for which there appear to be no reports of good outcomes and only reports of high failure rates, it seems clear that use of the Moje thumb CMC implant should be suspended until reports of good outcomes with adequate follow-up are published. For the Elektra implant, the only report of “good outcomes” was from the originator, who also reported a high rate of complications. Given the very poor results reported in this journal in 2012 by two different units, and no other independent reports of good outcomes, use of the implants should also be suspended until there are reports of good outcomes with adequate follow-up. The report in this edition is the first of the outcome of the Pi2 implant. The outcomes are no better than trapezectomy alone, and yet with a high rate of complications (six of 18 cases) (Maru et al., 2012). It would not seem justifiable to suspend use of the Pi2 implant following a single report, but surgeons should, in future, only use the Pi2 implant with caution. It is well-recognized that poor results are often not reported. Only following reports of poor outcomes by courageous surgeons do others start to voice their concerns. The recent experience in the UK of use of the LPM (Leuwen, Poeschmann, Metal) proximal interphalangeal joint arthroplasty clearly illustrates this (Field, 2008; Hobby et al., 2008).

Even if use of these implants is not suspended, it will be difficult to persuade patients to have them inserted, if they are consented appropriately, particularly with knowledge of the reported high failure rates. To rely on patient choice based upon informed consent is no longer appropriate for the Moje and Elektra implants. This is putting the responsibility onto our patients when we should make a stand as a profession and stop using implants with known poor outcomes unless other data are published to change our minds. Moreover, we should continue to be careful about being encouraged into using other new implants until adequate long-term follow-up is available.

Finally, on a more positive note I am pleased to report that the impact factor for this journal is now 1.17. This is due to the efforts of many people, particularly the authors, reviewers, and editors, and especially my predecessor as Editor-in-Chief, Professor Dias. I would like to thank everyone who has contributed and encourage you to submit even more good papers to raise the impact of the Journal even further.