A systematic review about treatment of osteoarthritis of the trapeziometacarpal joint with silicone implants and total joint replacement was performed. Good subjective outcomes were reported with Swanson prostheses, but their use cannot be recommended because of the risk of dislocation, implant failure, and foreign body reaction. With some of the more recent total joint arthroplasties, good medium-term survival rates were reported and better function in comparison with trapeziectomy and tendon arthroplasty, but prospective randomized studies with strong evidence were not available.
Key wordsTrapeziometacarpal osteoarthritis – trapeziometacarpal joint replacement – thumb carpometacarpal arthroplasty – survivorship
18 Silicone Implants and Total Joint Prostheses for Osteoarthritis of the Trapeziometacarpal Joint: A Systematic Review
In 1968 Swanson published the first report of a thumb carpometacarpal (CMC) joint arthroplasty using a silicone implant. 1 In 1973 de la Caffinière reported a technique with a total joint arthroplasty. 2 Since then many different designs have been developed.
A systematic review was performed to assess the results of thumb trapeziometacarpal arthroplasty. It followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline but was not registered in advance as it built upon a previous review. 3
18.2 Materials and Methods
The inclusion criteria for the systematic review were: full reports written in English or French of total joint replacement or silicone implants in patients with primary osteoarthritis of the trapeziometacarpal joint; studies reporting pain, strength, radiological changes, or failure rates as outcomes with a mean follow-up of at least 12 months; and studies with a design classification of levels I to VI, as classified by Jovell and Navarro-Rubio. 4
The exclusion criteria were: studies with a mean follow-up of less than 12 months; reports of patients with trapeziometacarpal joint osteoarthritis that could not be separated out from cases of secondary osteoarthritis (posttraumatic, rheumatoid, gout, chondrocalcinosis, or other inflammatory conditions); studies describing only surgical techniques, case reports, short report letters, revision procedures, cadaver studies, biomechanical studies, implants other than total joint arthroplasty or silicone implants such as tendon, ceramic, pyrocarbon, or interposition arthroplasty with synthetic or biological implants; and reviews and studies in which more than one type of implants were used and results were not reported separately.
The survival rates of implants and the method to determine these were recorded. Only articles with removal or revision of the prosthesis as end point were selected to assess survival. It was noted whether a cumulative survivorship analysis was performed and with which method. Usually the product-limit method 5 or the method of Armitage 6 was used. Implants were considered to have a good result if the failure rate was lower than 1% per year. For failure rate lower than 1% per year, follow-up of at least 10 years should be considered before any final conclusion is made. In cases of higher failure rates, results can be reported with shorter follow-up. 7
Assessment of the level of evidence was done according to the Jovell and Navarro-Rubio methodology 4 (Table 18.1) and assessment of the methodological quality with the Coleman methodology scoring system, part A 8 (Fig. 18.2).
Authors searched PubMed/Medline databases up to 31 January 2019. The following key terms were used for searching strategy: trapeziometacarpal osteoarthritis, trapeziometacarpal arthroplasty, trapeziometacarpal joint, trapeziometacarpal replacement, thumb CMC arthroplasty, thumb CMC joint replacement, and thumb CMC joint prosthesis. The bibliographies of all relevant papers and reviews were hand-searched.
A flowchart illustrating the selection of trials included in the systematic review is shown in Fig. 18.1.
18.3.1 Silicone Implants
Reports of 12 different silicone implants were found during the search. The first four were Swanson prostheses: (1) silicone trapezium implant with convex base, 1 (2) silicone trapezium implant with concave base 9 (Fig. 18.2), (3) silicone convex condylar implant, 10 (4) silicone concave condylar implant. 11 The first Swanson implants were constructed of common silicone elastomer; after 1974 high-performance elastomer was used. 11 Reports of other silicone implants found during the search include: (5) Kessler implant, a silicone implant with Dacron reinforcement. It is implanted without resection of the trapezium, 12 (6) a silicone sponge or disc for interposition in the trapeziometacarpal joint, 13 (7) the Ashworth-Blatt implant, a flat implant with a short stem that is inserted in the trapezium, 14 (8) the Niebauer arthroplasty, a Dacron-reinforced silicone trapezium implant, 15 (9) silicone implant with a tunnel in which a tendon strip can be placed to stabilize the implant, 16 (10) the Helal prosthesis, a silicone rubber ball with two stems, 17 (11) proplast silicone rubber trapezium implant without stem, 18 and (12) the tendon tie-in implant with a narrowed waist in which a tendon sling can be placed to stabilize the implant. 19
Twenty-two articles met the inclusion criteria with seven different types of implants. Only articles of one type of Swanson prosthesis were included, i.e., the Swanson trapezium silicone prosthesis with a concave base. Articles of the condylar Eaton implant 16 , 20 and the proplast silicone implant 18 were excluded. Results of Swanson prostheses are reported in Table 18.3 and Table 18.4 and of other silicone implants in Table 18.5 and Table 18.6.
Of the 12 articles about the Swanson prosthesis, 7 were evidence level VI, 4 level V, and 1 level III. The mean Coleman score was 34 (range: 17–50). The most commonly reported complications were dislocation and silicone synovitis. Reported failure rates ranged between 0 and 27%. The ten articles about other types of silicone implants were all retrospective with evidence level VI and only reported on a small number of patients. The mean Coleman score was 33 (range: 24–39). Reported complications were dislocation, foreign body reaction, and implant breakage. The highest failure rate was found with the Asworth-Blatt implant (55%).
18.3.2 Total Trapeziometacarpal Joint Implants
Results were found for 21 different total joint prostheses: Arpe prosthesis 40 (Fig. 18.3); Braun-Cutter prosthesis 41 , 42 , 43 ; Beznoska 44 ; Bichat prosthesis 45 ; cemented resurfacement arthroplasty 46 , 47 , 48 ; de la Caffinière prosthesis 49 ; Electra prosthesis 50 ; Guepar I prosthesis 51 , 52 , 53 ; Guepar II prosthesis 54 ; Isis 55 ; Ivory 56 (Fig. 18.4); Ledoux prosthesis 57 ; Maia 58 , 59 ; Mayo Clinic prosthesis 60 ; Moje Acamo 61 ; Moovis 62 ; Motec 63 ; Nahigian prosthesis 64 ; Roseland prosthesis 65 (Fig. 18.5); Rubis II 66 ; and Steffee prosthesis. 67
Papers concerning seven of these did not fit the eligibility criteria because either they included patients with conditions other than primary osteoarthritis (Braun, Guepar I, Ledoux, Maia, Mayo Clinic, and Steffee prostheses) or the results were published in Czech (Beznoska).
Forty-nine articles of the following total joint arthroplasties were included: Arpe (8), Braun-Cutter (1), cemented resurfacement (2), de la Caffinière (8), Electra (7), Guepar II (2), Isis (1), Ivory (4), Maia (3), Moje Acamo (3), Moovis (1), Motec (2), Nahigian (1), Roseland (5) (Fig. 18.5), and Rubis II (2). In one article results were reported of Electra and Motec prostheses. 63
Survivorship with removal of the prosthesis as an end point was reported on ten different total joint prostheses in 21 articles (Table 18.7). The most frequent were of the Arpe, de la Caffinière, and Electra prostheses. Results of those prostheses are presented in Table 18.8, Table 18.9, Table 18.10, Table 18.11, Table 18.12, and Table 18.13; results of other prostheses are shown in Table 18.14 and Table 18.15.