Trapeziectomy versus Maïa Prosthesis in Trapeziometacarpal Osteoarthritis

Abstract

Background  Osteoarthritis at the base of the thumb is the most frequent osteoarthritis of the hand. Trapeziectomy in a broad variety of surgical methods have been proposed to achieve pain reduction and improvement of thumb function. A well-known disadvantage is the long recovery time. Arthroplasty of the thumb carpometacarpal joint is a competing new method for this indication with different revision and complication rates reported.

Purposes  The aim of this study is to assess whether there are significant differences in outcome during the first 12 months and time return to work after either, implant of a Maïa joint prosthesis, or trapeziectomy with tendon interposition after Weilby.

Patients and Methods  This clinical follow-up study compares the efficacy of total basal joint replacement using the Maïa prosthesis with tendon interposition arthroplasty in 59 thumbs. Clinical, functional, and radiological results at preoperative, 3-, 6-, and 12-month postoperative are presented.

Results  We found a significant shorter return to work in the prosthesis group with 4.5 compared with 8.6 weeks. In addition to a significant difference in pain reduction with a better Mayo wrist score in the Maïa group after 3 months. The scores are closer after 6 months and nearly match after 12 months. Measurement of the pinch grip showed a parallel course. A radiological loosening of the cup in two patients was detected after 12 months.

Conclusion  Implantation of Maïa prosthesis enables a significant shorter recovery but is associated with the risk of loosening and higher costs.

Level of Evidence  Level IV, case–control study.

Osteoarthritis at the base of the thumb is a common problem and with 10% the most frequent osteoarthritis of the hand. It is caused by the special anatomy of the joint, which not only offers a wide range of motion but also induces a big strain to the joint of up to 200 kg in pinch grip. ^{1 2 3 4} Moderate to severe degrees of osteoarthritis might require surgical treatment for the purpose of restoring pain-free, strong, stable, and mobile function of the thumb and to improve patient global assessment. ⁴

Trapeziectomy remains the most common surgical treatment for thumb carpometacarpal (CMC) joint osteoarthritis. A broad variety of surgical methods, including simple trapeziectomy and trapeziectomy with interposition arthroplasty and/or ligament reconstruction, have been proposed to achieve all of these goals with a relatively long recovery time up to 12 months. ^{4 5 6 7 8} None of these surgical methods have proved to provide superior results compared with the other techniques. ¹ Furthermore, some of the procedures may be associated with negative effects such as deep pain and tendon problems. ^{9 10 11}

In 1973, De la Caffiniere described the replacement of the thumb CMC joint for the first time. ¹² Since this time, arthroplasty of this joint has become an established treatment. ² Many designs of cemented joint prosthesis have emerged, unfortunately associated with problems like aseptic loosening. ^{13 14 15} In recent years, new technologies have been applied to the design of uncemented trapeziometacarpal joint prosthesis. For better osteointegration, implant surfaces are coated with bone conductive materials, such as hydroxy apatite. In different studies partly, good long-term survival is reported. ^{1 6 16 17} The reported revision and complication rates differ in a high range. ^{2 4 6 18 19 20 21} In the German-speaking world, the use of prostheses in the treatment of thumb CMC osteoarthritis is only 5%, perhaps also caused by the higher costs of the prosthesis compared with the low-cost resection arthroplasty. ¹

The patients are becoming younger, and some are still employed when they need operative treatment. They not only want to return to work as soon as possible, but also to practice their prior sports after operation.

The aim of this study is to assess whether there are significant differences in time return to work and outcome during the first 12 months after either, implant of a Maïa joint prosthesis, or trapeziectomy with tendon interposition after Weilby. Further if there are differences in time trends, grip strength, pain, function, and complications after these procedures.

Patients and Methods

Clinical Series

This investigation was designed as a single-center retrospective study after obtaining approval of the institutional ethical committee.

The inclusion criteria were isolated trapeziometacarpal osteoarthritis determined by Eaton–Littler classification (stages 2–4) with osteoarthritic joint pain and loss of function. ²² All patients had at least 6 months of ineffective conservative treatment with hand therapy, supply with splinting, and oral pain therapy before operative treatment.

The included patients underwent surgery in the period from January 2018 to January 2021 at our hand unit with 12-month postoperative follow-up, including consequent aftercare at 3, 6, and 12 months.

Preoperatively, all patients were extensively informed about the two operative techniques and they decided about the method by themselves. Patients, who wanted implantation of a prosthesis underwent a computed tomography (CT) scan additionally to the routine radiographic investigation. Fifty-eight patients were eligible to be included in this study: 33 in the prothesis group and 25 in the Weilby group. These are all patients who have been treated in our institution because of thumb CMC osteoarthritis operatively. There were no patients excluded.

In the prosthesis group, there were 34 thumbs of 33 patients, with 19 left and 15 right hands compared with 11 left and 14 right thumbs in the Weilby group. The patients in the resection arthroplasty group had a median age of 64 years, compared with 59 years who received a Maïa implant. Overall demographic characteristics were similar between the groups ( Table 1 ).

Table 1. Patient data.

	Arthroplasty	Prosthesis	p -Value
Number of patients included	25	33
Number of thumbs included	25	34
Age (y)	64.12 (51–84)	59.25 (41–87)
Gender (female/male)	18 /7	25/9
Side (left/right)	11/14	19/15	0.256
Dominant hand (yes/no)	15/10	27/7	0.5
Eaton–Littler osteoarthritis stage	3.2 (2–4)	2.65 (2–3)	0.5
Occupation (employed/retired)	14/11	21/13	0.13

Surgical Technique

Joint Prosthesis Surgery

The Maïa TMC prosthesis (Group Lepin, Lyon Nord, France) is modular, cementless and hydroxyapatite-covered. The implant neck is composed of high-nitrogen-coated steel and the metacarpal stem and cup are composed of titanium alloy. Both components have a double layer coating consisting of a hydroxyapatite upper layer and a porous titanium lower layer that provides maximal secondary stability and bone ingrowth. The anatomical shape of the stem follows the profile of the first metacarpal (MC I). Four sizes of metacarpal stems are available (7, 8, 9, and 10 mm). The spherical metal backup is designed with a flat dome and thick rim to reinforce the press fit affect with a polyethylene liner. Four pins stabilize the cup. Two types of trapezial cups with two sizes (9 or 10 mm) are available. A range of modular offset necks reproduce the anatomical nonalignment of the first metacarpal with the center of the trapezium. Two series of removable necks (straight and offset) with three different lengths are available to provide maximal stability and avoid impingement.

The operations were performed under regional anesthesia with tourniquet and perioperative ceftriaxone use for infection prophylaxis. We use the dorsolateral approach with care to the tendons as well as the sensory nerve branches, followed by the winglike preparation of the joint capsule, which is sutured again after complete implantation. The resection of the base of the first metacarpal is done using a saw guide to determine the right angel and is aligned to the axis of the metacarpal. The resection should be at least 3 mm especially in cases with high deformity. To avoid leverage and restriction of range of motion, exophytes must be excised as far as possible from the MC I base and the trapezoid. Ligament repair at the CMC joint is not necessary after replacement.

The technique of implantation followed last surgical guidelines of Group Lepin to minimize complications after Maïa arthroplasty ( Fig. 1 ). ^{6 23}

Fig. 1.

Implantation of a MAÏA trapeziometacarpal joint prosthesis. ( A ) The line marks the planned skin incision. ( B ) After incision of the joint capsule and resection of the base of first metacarpal. ( C ) Searching the ideal position for the cup using a K-wire and checking by fluoroscopy. ( D ) Positioning the cup in the trapezoid without overlay. ( E ) Reaming the first metacarpal before inset of the stem. ( F ) After inset of the stem and neck with duo cup, prosthesis in situ with only mild tension.

The joints were immobilized in a forearm cast with the thumb included for 3 weeks, after which unlimited active motion was permitted. Simultaneously active and passive motion therapy was started in our hand therapy unit.

Tendon Interposition Arthroplasty

The surgical technique has been described by Weilby. ^{7 24}

Operations were also performed under regional anesthesia with use of tourniquet.

The thumb was immobilized in a forearm cast with thumb inclusion for 4 weeks, after which hand therapy was initiated with active and passive motion. Implantation of the prosthesis was performed by one surgeon (Dr. Johann Neureiter) and the resection arthroplasties were performed by three other experienced surgeons.

Assessment

Clinical and radiological assessment was performed preoperatively, at 3, 6, and 12 months postoperatively. Clinical examination consisted of key pinch strength measurement (Baseline Mechanical Pinch Gauge, 30 lb, Fabrication Enterprises, White Plains, NY) and evaluation of pain, using pain scale of Alnot/Saint Laurent (0–4). ²⁵ Patient satisfaction was measured using the modified Mayo wrist score questionnaire. Time of return to work was evaluated at the end of the study retrospectively.

The radiographic examination consisted of posterior–anterior and oblique radiographs. Preoperative radiographs were classified by Eaton–Littler, ²² and postoperative radiographs were assessed regarding implant alignment and loosening ( Fig. 2 ). Preoperative CT examination was performed in all patients of the prosthesis group for exact evaluation of the CMC joint concerning height of the trapezoid to determine whether cup implantation was possible. A minimal height of 8 mm was requested to fix the cup with an adequate bone stock remaining.

Fig. 2.

( A and B ) Posterior–anterior and oblique radiographs of a 65-year-old female patient. ( C and D ) Postoperative fluoroscopy after implantation of a MAÏA prosthesis. ( E and F ) Posterior–anterior and oblique radiograph of the same patient at 12-month follow-up. ( G and H ) Posterior–anterior and oblique radiographs of a 62-year-old female patient. ( I and J ) Postoperative fluoroscopy after resection arthroplasty. ( K and L ) Posterior–anterior and oblique radiograph of the same patient at 12-month follow-up.

In patients, who decided for implantation of a Maïa prosthesis, all contraindications for the implants had to be excluded (mild scaphotrapeziotrapezoidal [STT] osteoarthritis was tolerated), otherwise they received a resection arthroplasty. ¹

Operation time was collected as well as possible complications in the postoperative course and during follow-up period.

Statistical Analysis

For assessing the statistical significance of group-time interactions (i.e., to evaluate differences in time trends between groups), a nonparametric analysis of variance–type repeated measures test was used. ²⁶ To compare the outcomes between the two intervention groups at particular time points (i.e., 3, 6, and 12 months), the Wilcoxon–Mann–Whitney test was used. To control for multiple testing, all p -values were adjusted using the Bonferroni–Holm method. All statistical analyses were conducted using the statistical software R version 4.0.2. (R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria; 2020. https://www.R-project.org/ .)

Results

The patients in the resection arthroplasty group had a higher Eaton–Littler classification with 3.2 compared with 2.65, although the initial values of key pinch strength, pain scale, and Mayo wrist score have been nearly equal. ( Fig. 3A–C ).

Fig. 3.

( A ) Pinch grip (in kg), preoperative, 3- and 6-month postoperative with standard deviation, showing nearly similar courses. After 12 months significant stronger pinch grip in the Weilby group. ( B ) Course of pain (Alnot–Saint Laurent, 0–4, 0 is no pain, 4 is maximal pain) showing significant better pain reduction in the prosthesis group after 3 months and similar values after 6 and 12 months. ( C ) Course of Mayo wrist score showing nearly the same initial values and significant better results in the prosthesis group after 3 months and comparable results after 6 and 12 months (below 60: poor, 60–80: satisfactory, 80–90: good, 90–100 excellent).

Time required for implantation of a Maïa prosthesis was in median 60 minutes (40–105) compared with median 57 minutes (38–104) with no significant discrepancy.

Results of the key pinch strength are shown in Fig. 3A and pointed no significant difference after 3 and 6 months with significantly better results in the trapeziectomy cohort after 1 year. We found a difference in pain 3 months after operation, with significant benefit for the prosthesis group and almost equal after 6 and 12 months, shown in Fig. 3B . The modified Mayo wrist score showed a significant faster recovery of patients treated with prosthesis after 3 months but better results for the resection group at 6 and 12 months, where we found good subjective outcome (80–90) in both groups ( Fig. 3C ). Our findings show that there is a significant difference in time trends between the two groups, showing that the prosthesis group recovers faster in the first 3 months but thereafter makes limited progress, whereas the resection group shows a constant progress in function. Exact results of the different time points with p -values and relative effects are documented in Table 2 .

Table 2. Results.

	3 mo		6 mo		12 mo
	p -Value	RE (95% CI)	p -Value	RE (95% CI)	p -Value	RE (95% CI)
Pain	<0.001	0.95 (0.90–0.98)	0.623	0.58 (0.45–0.69)	0.338	0.62 (0.49–0.73)
Pinch grip	1.0	0.52 (0.37–0.67)	1.0	0.49 (0.35–0.64)	<0.001	0.85 (0.72–0.93)
Mayo	<0.001	0.08 (0.03–0.18)	0.026	0.71 (0.55–0.84)	0.338	0.63 (0.47–0.76)

Abbreviations: CI, confidence interval; Mayo, Mayo wrist score; Pain, pain measured according to the Alnot/Saint Laurent classification (0–4); Pinch grip, pinch grip in kg; RE, relative effect.

Notes: The relative effect quantifies the probability that the “trapeziectomy” group has higher scores on the outcome than the “prosthesis” group. For example, RE = 0.95 at the 3-month visit indicates that the estimated probability of an individual undergoing trapeziectomy to have a higher pain score than a subject with a prosthesis is 95%.

A significant difference could be found in the time return to work, with 4.5 (standard deviation [SD] = 0.9) weeks in the prosthesis group compared with 8.6 (SD = 2.3) weeks after trapeziectomy ( Fig. 4 ).

Fig. 4.

Time return to work in weeks. Significant shorter recovery time after implantation of a MAÏA prosthesis ( p  = 0.001).

Radiographic examination after 3 and 6 months demonstrated no signs of loosening of implants. At 1-year examination we found loosening of the cup in two patients without clinical deterioration; they showed no worsening of pain, grip strength, or Mayo score.

Preoperative CT scans discovered relevant information in eleven of the 34 patients with bony cysts in the trapezium or STT osteoarthritis, not seen in the radiographs ( Fig. 5 ). Formation and exact location of the cysts could be checked and gave important inputs for planning the operation, especially positioning of the cup. In two cases we found big cysts in which the surgeon decided to use a bone graft for adequate fixation of the cup, but one of them showed loosening after 12 months. The height of the trapezium was evaluated as adequate for cup implantation in all cases. Mild STT osteoarthritis was detected in three thumbs, but postoperative development of pain was not affected negatively in these patients. In two patients planed prothesis was changed to Trapeziectomy because of cyst formation combined with STT osteoarthritis.

Fig. 5.

( A and B ) Preoperative posterior–anterior and oblique radiograph of a patient before implantation of a MAÏA prosthesis (Eaton–Littler 3). ( C and D ) Computed tomography scan of the same patient showing a cyst in the trapezoid, not detected in the X-ray.

We found no significant peri- or postoperative complications in this series, especially no necessary operative revisions till 12 months after initial treatment.

Discussion

The results of our study demonstrate that patients treated with Maïa prosthesis because of trapeziometacarpal joint osteoarthritis of the thumb, can return significantly earlier to work (4.5–8.6 wk, p  = 0.001) with significant earlier pain reduction ( p  = 0.001) and better Mayo wrist score after 3 months than patients treated with trapeziectomy and tendon interposition after Weilby. In contrast the statistical time trend differed significantly with early fast recovery for the prosthesis and flattening of the improvement after 3 months, compared with the constant improvement in the Weilby group ( p  < 0.0001). They even showed a significant better result after 12 months in pinch grip ( p  = 0.001), despite slightly inferior basic values. During 1-year follow-up no revision surgery was necessary, standard X-ray showed two patients with cup loosening after 12 months but clinical inconspicuous function.

Resection arthroplasty with tendon interposition after Weilby is well known and results of large trials are published with good long-term outcome, but recovery of grip strength, range of motion, pain reduction, and function need time. ^{1 2 6 16 27} Vermeulen and Vadstrup published nearly similar results to our Weilby group. Rehabilitation made the major advance between 3 and 6 months after operation with only a slight increase in function from 6 to 12 months. After 1-year, acceptable hand function with good use in daily living was achieved with persistent good function subsequently. In contrast to the findings of Vadstrup, who describes 10% of their patients without pain reduction, all our patients showed an acceptable pain reduction combined with good subjective results documented by the course of Mayo wrist score. ^{7 8} Because of young patients with growing demands different approaches to address the time problem have been searched for. Since the results of De la Caffiniere in 1973, ¹² the quality and function of implants showed a great enhancement, current products use coated surfaces for better osteointegration, variable geometries, and long-lasting material combinations for better survival rates. ^{1 6 16 17}

Despite the huge stress to the joint, studies with a follow-up of 10 years show satisfactory survival rates between 73 and 93%. ^{2 6 28} The partially high rate of reported revisions up to 44% is considerable ^{18 19 29} in contrast to only 4.6% reported by Martin-Ferrero, who showed radiological loosening only in 7.7% after 10 years. ² In our series we have noted no necessary revisions within 12 months, but two patients with radiographic signs of loosening of the cup (6%) without functional impairment. In case of necessary revision, acceptable functional results of secondary resection arthroplasty are published comparable to these of primary arthroplasties. ^{30 31}

Operative revisions are rare after resection arthroplasty including impingement, nerve lesions, and tendinopathy. ^{32 33}

One of two patients with cup loosening showed a huge cyst in the trapezoid in the preoperative CT and needed bone grafting for fixing the cup. Incidentally, we would not perform this again and recommend a resection arthroplasty, because of the risk of further progression of cup loosening.

Because of the findings in the CT scans with therapy relevant information, we will continue this procedure in future. Mild variants of STT osteoarthritis, observed in three patients, didn't develop pain after implantation of prosthesis. Important seems to avoid high tension in implantation of the prosthesis like recommended in the current surgical guidelines. ²³ So, for us, it isn't still a contraindication for a prosthesis.

In the literature we found only two papers with a similar study design, Jager et al, who has a comparable cohort, analyzed the course of pain reduction, pinch grip, and functional scores and described a significant better outcome after 6 months for the prosthesis, also using Maïa prosthesis. They reasoned that implants should be recommended to patients because of the earlier recovery. ¹⁷ In contrast, we found nearly comparable results in our two groups after 6 months and even better results in the Weilby group after 1 year.

Ulrich-Vinther et al, compared tendon interposition arthroplasty with Electra prosthesis over 12 months. ⁴ The age and the follow-up of the patients are similar to our cohort. Concerning the initial Eaton–Littler stage, the resection arthroplasty group was with a median of 2.2 much better than ours with 3.2, with similar values in the prosthesis groups. Although the worse Eaton–Littler score preoperatively, only the key pinch grip was inferior in the Weilby group and modified Mayo wrist score and pain scale have been comparable before operation. So, we agree with the authors, who found also two signs of osteolysis without clinical relevance, that the prothesis enables a faster recovery after operation with comparably low rates of complications. However, the described persistent drawback of the arthroplasty group after 12 months does not concur to our results.

The objective fact of shorter time return to work for us is the most important justification to implant a prosthesis with costs of 1970 €, alongside the subjective enhancement in life quality in the first 3 months after operation.

Although our results are convincing, there are some limitations of this study. We analyzed the data retrospectively without random assignment to the operative method. The patients in the Weilby group are older, and there are some patients with Eaton–Littler IV, which were not present in the prosthesis group. And while we could show a significant faster return to work and earlier pain reduction with better modified Mayo wrist score after 3 months, we cannot make any comment about function and possible complications after 5 or 10 years for the Maïa group.

Considering our results implantation of a Maïa prosthesis can be recommended to patients who need a fast recovery and early return to work as possible, but they have to be informed about the risk of cup loosening or migration with probably necessary revision surgery. Mild STT osteoarthritis seems to be no contraindication if avoiding high tension in implantation.

Funding Statement

Funding This investigation was done in the AUVA Trauma Center Salzburg, Austria and the institution received no funding or financial support for this investigation.