Dual mobility prosthesis for trapeziometacarpal osteoarthritis: results from a prospective study of 55 prostheses

Introduction

Osteoarthritis of the trapeziometacarpal (TMC) joint is a common disease in the elderly and particularly menopausal women (Armstrong et al., 1994). If patients do not respond to a prolonged comprehensive conservative treatment, including splinting, physiotherapy, anti-inflammatory drugs and change of lifestyle, surgical procedures can be considered. The standard procedure is simple trapeziectomy with ligament reconstruction (Hamasaki et al., 2021; Yeoman et al., 2019). However, trapeziectomy can be associated with reduction of thumb length due to proximal migration of the first metacarpal bone, which may lead to loss of grip and pinch strength (Yang and Weiland, 1998). TMC joint arthroplasty is a trapezium-preserving alternative. It was devised to replace the joint and restore normal thumb length in order to improve thumb grip strength, stability and motion. Most TMC joint implants available on the market had a single mobility design. However, implant loosening and dislocation of the prosthesis were frequently observed complications in this type of arthroplasty (Andrzejewski and Ledoux, 2019; Dehl et al., 2017; Froschauer et al., 2020; Tchurukdichian et al., 2020).

To improve stability of the prosthesis, while preserving the wide TMC joint range of motion (ROM), the dual mobility principle based on hip arthroplasty was introduced (Farizon et al., 1998). These prostheses consist of a double-sided concentric articulation with a metal–polyethylene bearing. The two articulating surfaces give the thumb prosthesis more stability and up to a 117° of ROM (Brauns et al., 2019). However, few studies have been reporting results of dual mobility protheses (Lussiez et al., 2017, 2021). Therefore, the aim of this prospective study was to examine functional and radiographic outcomes as well as correction of thumb metacarpophalangeal (MCP) joint hyperextension after total TMC joint replacement with a dual mobility prosthesis. We hypothesized that correction of thumb length may also improve MCP joint hyperextension.

Methods

After institutional review board approval (2019-14184A), a prospective study was conducted at two centres (BG Trauma Center Ludwigshafen and University of Heidelberg) to investigate functional results of patients who received TMC arthroplasty with a dual mobility prosthesis (TOUCH®, KeriMedical, Les Acacias, Switzerland) for surgical treatment of TMC joint osteoarthritis. Surgical complications as well as clinical and radiological outcomes were evaluated. Results were analysed and compared with preoperative measurements. In addition, outcomes were also compared with the contralateral thumb (non-operated and no clinical sign of TMC joint osteoarthritis), which served as a healthy control in cases of unilateral disease.

From July 2018 to January 2022, a total of 55 TMC prostheses (52 patients) with a minimum follow-up of 1 year were included in this study. All patients suffered from primary degenerative osteoarthritis and had ceased responding to conservative treatment. The criteria for total joint prosthesis implantation were painful primary osteoarthritis (limited to Eaton and Littler Stage II or III), a trapezium height of at least 9 mm, and a patient profile of moderate physical load of the hand (defined as not working with heavy or vibrating machines). Prior to surgery, each patient was informed about trapeziectomy as a surgical alternative to the dual mobility arthroplasty. Patients were free to choose their preferred surgical treatment. Exclusion criteria were Stage I and IV osteoarthritis or refusal to participate in the study. Patients with Stage I were offered conservative treatment. Patients with Stage IV received alternative operative treatments also addressing the scaphotrapeziotrapezoidal arthritis, if symptomatic. Patients’ characteristics were collected during informed consent for surgery.

Patients included in the study were assessed pre- and postoperatively at 1.5, 3, 6 and 12 months and then continued yearly. Three experienced Level 4 senior surgeons performed all operations (Tang and Giddins, 2016). The outcomes were assessed independently of the surgeons by the first two authors (FF and AMT). Pain reduction was evaluated using a visual analogue scale (VAS) from 0 (no pain) to 10 (worst pain imaginable), when performing exercises such as ‘opening a bottle cap’ or ‘door opening with a key’. Follow-up examinations included ROM measurements of the thumb MCP joint using a standard goniometer. To assess radial and palmar abduction, the angle between the thumb and index finger metacarpal was measured. Thumb opposition was evaluated using the Kapandji score (Goubau et al., 2013). Grip and key pinch strength were assessed by the Jamar dynamometer and pinch meter (Jamar Technologies Inc. Clifton, NJ, USA). During follow-up, surgical complications such as aseptic loosing or dislocation were recorded. Each patient completed the Quick Disabilities of the Arm, Shoulder and Hand (QuickDASH) (Beaton et al., 2005), the Michigan Hand Questionnaire (MHQ) (Chung et al., 1998) and the 36-Item Short Form Survey (SF-36) for quality of life evaluation (Ware and Sherbourne, 1992). In addition, patients were asked if they would undergo this surgery again.

Radiographs of the thumb ray were taken in posterioanterior and lateral views preoperatively (Figure 1), immediately after operation, and at each postoperative visit. Radiographs were evaluated for radiolucent zones in the trapezium or metacarpal and for migration or dislocation of the prosthesis by LH, FF, AMT. To examine postoperative thumb length restoration, the ratio between first metacarpal–trapezium and second metacarpal–trapezoid length (M1/M2) was calculated on posteroanterior X-rays of the hand, as described by Ledoux (2017) with a normal M1/M2 ratio of 0.73 (SD 0.05). MCP joint hyperextension was measured on lateral view X-rays in resting position without pushing on something on DeepUnity X-ray viewer (Dedalus Healthcare Group AG, Bonn, Germany) and clinically with a finger goniometer by the second author (AMT).

OPEN IN VIEWER

Figure 1.

Pre- (a) and postoperative radiographs (b) of a 57-year-old woman at 12 months follow-up.

The Touch® dual mobility prosthesis is a modular, spherical and cementless prosthesis. The TMC joint was approached using a dorsoradial skin incision between the tendons of the abductor pollicis longus and extensor pollicis brevis (Figure S1A–L). After insertion of the prosthesis, meticulous suture of the capsuloperiosteal flap was performed. Postoperatively, all thumbs were immobilized in a thumb brace with the thumb metacarpal in slight radial abduction for 1 week, followed by active thumb mobilization. Full load bearing was allowed after 6 weeks.

Results

Fifty-two patients (55 prostheses) were included in the study (Figure 2). Seven patients had a painful TMC joint osteoarthritis affecting the contralateral thumb and were planned for prosthesis implantation in the future. Patient characteristics are shown in Table 1. Eight thumbs (15%) had Stage II osteoarthritis and 47 (85%) Stage III according to the Eaton and Littler classification. The mean postoperative follow-up was 25 months (range 12–36). Fifty thumbs (91%) reached the 2-year follow-up and no patients dropped out.

OPEN IN VIEWER

Figure 2.

Flowchart of patients included in the study.

OPEN IN VIEWER

Table 1.

Patient characteristics and comorbidities of patients with a Touch® prosthesis.

Patient details	Total
Patients	52
Thumb trapeziometacarpal joint arthroplasty	55
Age (years)	58 (42–76)
Gender (women/men)	38/14
Operative time (minutes)	76 (49–140)
Arthroplasty on dominant hand	22 (42%)
Working/retired	43 (83%)/9 (17%)
High manual strain/low manual strain	6 (12%)/46 (88%)
Length of postoperative hospital stay (days)	2.4 (1–5)
Comorbidities:
Arterial hypertension	7 (13%)
Chronic obstructive pulmonary disease	3 (6%)
Active smoker at time of surgery	3 (6%)
Coronary heart disease	2 (4%)
Body mass index	25 (18–32)

Data presented as number (%) or mean (range).

The mean VAS for pain had decreased significantly 1.5 months after surgery (mean difference: 5.1; 95% CI: 4.4 to 5.8; p < 0.0001), with pain reduction remaining statistically significant throughout the entire follow-up period (Table S1, Figure 3). Mean carpometacarpal joint movement showed a significant increase in radial and palmar abduction at 3 months postoperatively (mean difference: 5.3°; 95% CI: 2.7 to 5.7; p < 0.0001; mean difference 6.3°; 95% CI: 3.8 to 8.7; p < 0.0001) and continued to improve further. In addition, improvement of mean thumb opposition according to the Kapandji index already reached significance at 1.5 months after surgery (mean difference: 1.1; 95% CI: 0.7 to 1.5; p < 0.0001). Mean grip and pinch strength improved significantly 3 months postoperatively (mean difference 4.1 kg; 95% CI: 2.4 to 5.8; p < 0.0001; mean difference 1.5 kg; 95% CI: 0.9 to 2.1; p < 0.0001) (Figure 4).

OPEN IN VIEWER

Figure 3.

Graphs showing mean (error bars: SD), (a) visual analogue scale score for pain during active thumb movements, (b) total range of active extension and flexion of the metacarpophalangeal joint of the thumb, (c) active palmar abduction and (d) active radial abduction at the trapeziometacarpal joint compared with preoperative values.

OPEN IN VIEWER

Figure 4.

Graphs showing mean (error bars: SD), (a) Kapandji index, (b) grip- and (c) pinch strength with the evolution of time compared with preoperative values.

The mean QuickDASH of 14 (range 6–28) and MHQ scores 82 (range 67–92) revealed overall good functional results at 1 year after surgery (Figure 5). During follow-up, 50 of the 52 patients (96%) stated that they were happy with the outcomes of the procedure and would agree to it again, if given the choice. No statistically significant differences were found between the clinical outcomes of the operated side versus the clinically inapparent contralateral side at the 12 months follow-up (Table S2).

OPEN IN VIEWER

Figure 5.

Boxplot graphs (minimum, 25th percentile, median, 75th percentile, maximum) illustrate the comparison of preoperative and scores at 1 year after surgery of (a) the 36-Item Short Form Survey (SF-36) for quality of life, (b) the Michigan Hand Questionnaire (MHQ) QuickDASH and (c) scores.

Prior to operation, 18 thumbs (33%) with Eaton and Littler Stage III suffered from MCP joint hyperextension of more than 15° (mean 19°; range 15°–28°). In all of these patients, a significant normalization in the position of the MCP joint was observed as early as 1.5 months postoperatively (mean difference 12°; 95% CI: 8.9 to 14.6; p < 0.0001), which remained stable until the last follow-up examination (mean difference, range 0°–5°) without additional surgery to this joint (mean difference 17°; 95% CI: 14.5 to 19.4; p < 0.0001) (Figures 6 and 7).

OPEN IN VIEWER

Figure 6.

Preoperative and postoperative metacarpophalangeal (MCP) joint hyperextension before and after operation.

OPEN IN VIEWER

Figure 7.

(a) Pre- and (b) 12 months postoperative radiographs of a 63-year-old woman demonstrating normalization of pre-existing metacarpophalangeal joint hyperextension.

Postoperative radiographs revealed that the mean first metacarpal–trapezium length M1: 60 mm (SD 5) was 6.1% longer than the respective preoperative values M1: 56 mm (SD 4). Thus, an adequate restoration of thumb length was achieved. The postoperative M1/M2 ratios remained stable over the course of all follow-up evaluations (mean 0.77 to 0.78), thereby ruling out significant cup or stem sinking (Table S3). Radiographs did not show any sign of periprosthetic radiolucent zones. In two thumbs, a small bony cyst near the cup was detected, which did not affect prosthetic stability.

During follow-up examination, no revisions due to infection, loosening, dislocation or implant breakage were necessary.

Discussion

The dual mobility design has helped to achieve particularly favourable results regarding the ROM and early recovery of grip and pinch strengths. In 2021, Lussiez et al. were the first to publish their results of 107 dual mobility Touch® implantations at a minimal follow-up of 3 years. They reported a significant improvement of radial abduction (29° to 42°) and Kapandji scores (8 to 9). Mean pain intensity (VAS 1–10) had decreased from 7 to 1 and mean key-pinch strength had increased from 4 kg to 6 kg (Lussiez et al., 2021). These findings correspond with our results for improvement in mean TMC joint radial-abduction (38° to 48°), key-pinch strength (3 kg to 6 kg) and pain intensity (7 to 1) 3 years after surgery. However, Lussiez et al. (2021) did not report the early postoperative recovery, as patients were only examined annually and comparisons with contralateral healthy thumbs were not shown. In our cohort of patients with TMC joint osteoarthritis, significant improvements in radial and palmar abduction were already achieved 3 months after surgery. In accordance, thumb opposition, determined with the Kapandji score, significantly exceeded preoperative values as early as 1.5 months after surgery. We found both grip and pinch strengths had significantly improved at 3 months postoperatively, and increased significantly during the further follow-up. We conclude that during the early postoperative course, it can be expected that the grip strength increases 25% at 3 months and 33% at 12 months, while the pinch strength increases 40% at 3 months and 50% at 12 months after surgery. One year after surgery, we could demonstrate that values of TMC joint ROM and pinch strength were comparable with the contralateral side. This was underlined by QuickDASH score improvements at 1 year, which almost reached normal values, as suggested by Hunsaker et al. (2002). In this context, it is worth noting that 83% of our patients were manual workers at the time of surgery. Therefore, early postoperative recovery is important for return to everyday work.

Janakiramanan et al. (2021) prospectively investigated early functional recovery after trapeziectomy and tendon interposition in 55 patients up to 12 months post-surgery. Interestingly, the reported rehabilitation values for pain and ROM were similar to our results, whereas in our study with dual mobility prostheses, superior outcomes of key pinch strength at 3, 6 and 12 months were achieved. In this context, some authors suggest that a decrease of pinch strength may result from reduced thumb length after trapeziectomy (Downing and Davis, 2001). However, the influence of scaphometacarpal height on thumb weakness remains to be answered. Nevertheless, our study clearly demonstrated successful thumb length restoration according to the M1/M2 ratios, which were similar to the results shown by Ledoux (2017).

In our study, 18 thumbs presenting with MCP joint hyperextension exceeding 15° (mean 18°) were included. We demonstrated that a significant normalization of MCP joint hyperextension was reached as early as 1.5 months after surgery, which remained stable until the latest follow-up. Our findings correspond to those of Lussiez et al. (2021), who reported that MCP joint hyperextension was corrected in 19 patients by implanting the dual mobility Touch® prosthesis.

Lastly, we observed two small bony cysts in the trapezium at 6 to 12 months after surgery, which was not considered a complication, because the cups were still completely anchored in the trapezial cavity. In our opinion, these findings might have resulted either from a low bone density or from cavity preparation in the trapezium. Both patients had no clinical complaints. In comparison with other prostheses, the dual mobility Touch® prosthesis has a metal-on-polyethylene bearing configuration, which may decrease the risk for metal abrasion and loosing of the trapezial component over the time (Hansen et al., 2013; Huang et al., 2015). The dual mobility design includes a small and large joint reaching a total rotational ROM of 117°. The small joint cup minimizes the contact surface between the neck and the cup, which is thought to increase biomechanical stability. This is supported by the absence of dislocation in our cohort. Also single mobility prostheses have shown survival rates up to 95% after 10 years of follow-up (Tchurukdichian et al., 2020). For this reason, our follow-up period might be too short to draw definite conclusions about the benefits of the dual mobility design in terms of survival/dislocation rates compared with single mobility prostheses.

Despite our promising results, the study has several limitations that should be considered. First, our study did not reach mid-term nor long-term follow-up, as is recommended for prostheses (Tang et al., 2019). As described in the study design, the follow-up of the whole cohort will be continued with annual assessments to yield long-term results. Second, we did not compare our cohort of prostheses with a cohort of patients who fulfilled the prosthetic criteria but chose another surgical procedure. So far, a comparison of dual mobility prosthesis implantation to standard surgical procedures, such as trapeziectomy, is still lacking. Third, we compared our results with the contralateral clinically asymptomatic side, therefore, some degree of osteoarthritis may have been present in the controls.

Supplemental Material