Abstract
Purpose
Thumb carpometacarpal osteoarthritis (CMC OA) is a common condition associated with functional limitations and pain. Surgical interventions for CMC OA include simple trapeziectomy, ligament reconstruction and tendon interposition, arthrodesis, and joint replacement. The aim of this review was to evaluate the comparative effectiveness of surgical techniques.
Methods
A systematic review was performed, and the Medline, EMBASE, and SCOPUS databases were searched. Randomized controlled trials (RCTs) investigating the surgical management of CMC OA were included. The outcomes of interest were long-term pain reduction (visual analog scale), complications, and functional improvement (Disabilities of Arm, Shoulder and Hand questionnaire). Bayesian network meta-analyses were conducted, and longitudinal analysis of complication severity was performed using the Kruskal-Wallis H test.
Results
There were 26 RCTs included, representing 19 surgical techniques and 1,193 hands. The average follow-up period was 29 months. In the analyses of pain, simple trapeziectomy ranked first, and trapeziectomy with button ranked last. Uncemented joint replacement ranked first in the analyses of function. Trapeziectomy alone demonstrated the most favorable results regarding avoidance of complications, and arthrodesis with plate and screw ranked last in the analysis of complications. Alternative surgical techniques were associated with a relative increase in complication severity across all time periods (χ2(2) = 22.92, 46.82, and 7.01 after 0–3, 3–12, and 12+ months; P < .001, <.001, and .03, respectively).
Conclusions
Our findings demonstrate that simple trapeziectomy is effective at relieving the pain associated with CMC OA and minimizing postoperative complications, demonstrating a relative increase in both the number and severity of complications after alternative surgical techniques.
Clinical relevance
Our review supports the use of simple trapeziectomy as the mainstay of surgical management of CMC OA, providing the foundation for research investigating metacarpal subsidence, trapeziometacarpal biomechanics, and implant dynamics in patients with CMC OA.
Key words: Network meta-analysis, Osteoarthritis, Pain, Trapeziectomy
Thumb carpometacarpal osteoarthritis (CMC OA) is a common condition with substantial population-level morbidity.1 In addition to pain and functional limitations, the disability associated with CMC OA can have a considerable neuropsychological impact.2,3 Although conservative therapies for CMC OA include splinting, physiotherapy, and topical anti-inflammatory gels, these interventions may be ineffective in patients with severe degenerative disease.4 International guidelines including the European League Against Rheumatism recommend operative treatment after other modalities have failed.5 There is a wide range of surgical treatments available for CMC OA, and the most established surgical technique for CMC OA is trapeziectomy alone (simple trapeziectomy). Results from a cohort study of over 40,000 patients registered through the National Health Service demonstrated a very low complication rate after simple trapeziectomy, and recent health economic evaluation analyses have confirmed that simple trapeziectomy is more cost effective than nonsurgical treatment alone.6,7
Simple trapeziectomy may be augmented using ligament reconstruction and tendon interposition (LRTI). These concomitant soft tissue procedures can provide greater joint stability by preventing proximal displacement of the head of the metacarpal after trapeziectomy.8 However, the clinical significance of decreased scaphometacarpal distance is unclear, with clinical and cadaveric studies demonstrating little effect on pain and carpal stability.8,9 Alternative surgical techniques for CMC OA including hemitrapeziectomy, Poly-L/D-lactic acid (PLDLA) implant, joint replacement, and arthrodesis have also arisen.10 Overall, there remains a lack of robust evidence to guide the surgical management of CMC OA. The systematic review by Li et al11 reported equivocal results when comparing the efficacy of simple trapeziectomy with trapeziectomy and LRTI. Although Qureshi et al12 describe a relative improvement in Disability of Arm, Shoulder and Hand (DASH) and Kapandji score after trapeziometacarpal joint replacement compared to trapeziectomy with LRTI, this review also describes a higher complication rate after trapeziometacarpal joint replacement, ultimately highlighting the need for further research in order to compare treatment modalities.
In the context of the wide range of surgical techniques available for CMC OA combined with the relative lack evidence comparing techniques, we carried out a systematic review and network meta-analysis to investigate outcomes after surgical treatment of CMC OA. Network meta-analysis is a relatively new statistical technique that simultaneously evaluates multiple interventions using direct and indirect evidence. The purpose of this review was to evaluate the comparative effectiveness of different surgical techniques for CMC OA in improving postoperative outcomes, with the aim of evaluating the outcomes of pain, functional improvement, and complications.
Materials and Methods
This review was carried out in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (http://www.prisma-statement.org/) and registered on the International Prospective Register of Systematic Reviews (PROSPERO; registration number CRD42024533328).
The Medline, EMBASE, and SCOPUS databases were independently searched by two reviewers. The last search was conducted in August, 2024. Citation matching was performed to find additional eligible studies. To minimize the risk of publication bias, gray literature was searched via OpenGrey. The full search strategy is outlined in Appendix 1 (available on the Journal’s website at www.jhsgo.org).
Published randomized controlled trials with all study arms investigating surgical techniques for CMC OA were included. Studies investigating nonsurgical treatments (including physiotherapy, extracorporeal shock wave therapy, and intra-articular injection) or alternative medicine (prolotherapy, leech therapy, and acupuncture) were excluded. NonEnglish articles were excluded.
Risk of bias was assessed using the Cochrane tool, with a focus on the following constructs: random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), and selective reporting (reporting bias).13
The outcomes of interest were long-term pain, functional improvement, and complications.
Pain was measured using the visual analog scale (VAS), and long-term pain reduction was defined as VAS improvement greater than 6 months after surgery. Function was measured using the Disabilities of Arm, Shoulder, and Hand questionnaire (DASH). Based on discussions between both authors, complications were classified as mild (including superficial wound infection, mild sensory changes, and mild scar tenderness), moderate (including trigger thumb, tendinitis at site of harvested tendon, and reflex sympathetic dystrophy), and severe (reoperation, osteolysis of implant, device intolerance, neuroma requiring surgery, and nonunion requiring revision surgery). The follow-up period at which complications arose was also extracted and classified as immediate (within 3 months after surgery), intermediate (between 3 and 12 months), and delayed (more than 12 months).
Additionally, the following information was extracted from each study: author, year of publication, region, follow-up length, number of patients in each study arm, and surgical technique. Additional data about individual patient characteristics were not extracted from each study because of inconsistent reporting across included RCTs.
When articles failed to report the mean and SD, the median, range, interquartile range, and number of participants was used to obtain the mean and SD. This was performed according to Luo et al (2018)14 and Wan et al (2014).15
Details of data transformation are outlined in Appendix 2 (available on the Journal’s website at www.jhsgo.org).
The data were organized using Review Manager (RevMan) Version 5.3.16 Network meta-analysis was conducted using MetaInsight Version 4.1.0.17 This is an open-source program using RShiny and netmeta that allows for Bayesian network meta-analyses with real-time sensitivity analyses. The reference treatment was simple open trapeziectomy (using a radiodorsal incision and without LRTI). Separate network meta-analyses were created for pain and functional improvement, as indicated by the continuous outcome of mean difference in VAS or DASH score after surgery.
Surface under the cumulative ranking curve (SUCRA) plots were constructed to illustrate treatment hierarchies.18 This enabled graphical representation of treatment efficacy based on rankings obtained in the network meta-analysis. The SUCRA values obtained from the Bayesian network meta-analysis indicate the probability a particular surgical treatment is ranked highest for the outcome of interest, considering the inherent uncertainty introduced from heterogeneity within studies. Higher SUCRA values indicate a more favorable ranking, and lower values indicate a lower likelihood of treatment superiority.19 Ranking were based on smallest postoperative incidence of surgical complications, greatest mean improvement in VAS score after surgery (ie, improvement in pain), and greatest reduction in DASH score after surgery (ie, functional improvement).
Finally, in addition to the Bayesian analysis, complications were also analyzed based on severity (mild, moderate, severe) and follow-up period (immediate, intermediate, delayed).20 After testing for normality (Kolmogorov–Smirnov test), the complication data were ranked based on severity and analyzed using the Kruskal-Wallis H test. This is a nonparametric test that was used to identify significant differences in the distribution of complication severity after simple trapeziectomy, trapeziectomy with concomitant soft tissue procedure, and alternative surgical techniques (including arthroplasty and spacer interposition). Combined with the P value, the H statistic obtained from the Kruskal–Wallis test indicates the strength of evidence suggesting that the distribution of complication severity is significantly different across surgical techniques. A high H statistic indicates that the choice of surgical technique may have affected the severity of resultant complications, whereas a low H statistic suggests that distribution of complication severity is independent of choice of procedure. At a significance level of 0.05 (meaning that p values of less than 0.05 were considered significant) and as there were three groups included in the analysis (patients who had undergone simple trapeziectomy, LRTI, or alternative technique), the critical value of H was set at 5.99. Thus, H statistic values greater than 5.99 were high, and values below 5.99 were considered to be low. The Kruskal–Wallis H test was repeated for all three time periods to ascertain the severity of complications over time.
The statistical techniques used in this review were validated by an independent statistician.
Results
There were 26 studies included from the database search (Fig. 1), representing 19 different surgical techniques and 1193 operations. No additional studies were included from the gray literature search. The average follow-up period was approximately 2 years and 5 months. Trapeziectomy with LRTI with the Burton-Pellegrini technique was the most commonly investigated technique (14 studies), followed by simple trapeziectomy 12 studies.21 Supplemental Table S1 (available on the Journal’s website at www.jhsgo.org) outlines the characteristics of included studies. Supplemental Table S2 (available on the Journal’s website at www.jhsgo.org) outlines the outcomes measured in each study.
Figure 1.
Preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement of search results.∗
∗Flowchart illustrating the inclusion and exclusion process of studies identified from the literature search.
The most common bias found among included studies was performance bias because of incomplete blinding of participants and personnel (Supplemental Fig. S1, available on the Journal’s website at www.jhsgo.org). There were nine studies determined to be at high risk of performance bias. This is common among surgical research, as blinding is difficult to achieve. Most studies used an independent third party to assess outcomes after surgery, and six studies were deemed to be at high risk of detection bias. Adequate randomization was achieved in 24 of the 26 studies, with most using computer generated randomization. Most studies performed randomization immediately before surgery, thereby decreasing selection bias. Approximately one-third of the included studies were classified as high risk for attrition bias because of incomplete outcome data. Selective reporting bias was not identified in any of the included studies. Finally, publication bias was not formally assessed using statistical tests as a symmetrical distribution of effect sizes could not be assumed because of differences among included studies and surgical techniques.
A total of eight RCTs investigated VAS score after CMC OA surgery. Overall, simple trapeziectomy resulted in the greatest improvement in VAS score (mean difference: –6.86; SD: 2.45), followed by the Swanson silicone implant (mean difference: –5.9; SD: 1.78) (Table 1). Trapeziectomy with button resulted in the least reduction in postoperative VAS score (mean difference: –2.8; SD: 3.40).
Table 1.
Pooled Average∗ VAS Scores After CMC OA Surgery
| Technique | No. of Studies | Total No. of Hands | Average Mean Difference in VAS Score | SD |
|---|---|---|---|---|
| Simple trapeziectomy | 3 | 88 | –6.86 | 2.45 |
| Silicone implant (Swanson) | 1 | 13 | –5.9 | 1.78 |
| LRTI (APL) – Sigfusson and Lundborg | 5 | 178 | –4.96 | 1.65 |
| PLDLA implant (Artelon) | 3 | 64 | –2.60† | 1.02 |
| Trapeziectomy with collagen xenograft | 1 | 13 | –3.04 | 2.68 |
| LRTI (FCR) – Burton-Pellegrini | 2 | 51 | –2.83 | 1.04 |
| Trapeziectomy with button | 1 | 37 | –2.8 | 3.40 |
APL, abductor pollicus longus; FCR, flexor carpi radialis; LRTI, ligament reconstruction and tendon interposition; PLDLA, poly-d-lactic acid; VAS, Visual Analog Scale.
Unweighted mean improvement in VAS scores after each surgical technique reported across all studies.
Indicates disagreement between average mean difference across included studies and Bayesian network meta-analysis ranking.
Except for the PLDLA implant (Artelon spacer), the results of the Bayesian network meta-analysis ranking supported the results obtained from individual RCTs (Fig. 2). Simple trapeziectomy was ranked first (SUCRA value ∼80%), followed by the Swanson silicone implant (SUCRA value ∼73%), and trapeziectomy with button was ranked last (SUCRA value ∼20%).
Figure 2.
Surface under the cumulative ranking curve (SUCRA) diagram∗ for pain (mean difference) after each surgical technique. FCR, flexor carpi radialis; LRTI, ligament reconstruction and tendon interposition; PLDLA, poly-d-lactic acid.
∗Diagram illustrating the ranking and cumulative probability of rankings associated with each surgical technique. Simple trapeziectomy is ranked first, and trapeziectomy with suture-button suspensionplasty is ranked last.
Two studies failed to report complications, and one study failed to report the timing of specific complications. Therefore, a total of 22 studies were included in the analysis of complications (Table 2), representing 15 surgical techniques.
Table 2.
Summary of Complications Reported Across All Studies
| Severity∗ | Complication | Associated Surgical Technique |
|---|---|---|
| Mild | Superficial surgical site infection | Simple trapeziectomy, trapeziectomy with button, trapeziectomy with LRTI |
| Hematoma around tendon harvesting site | Trapeziectomy with LRTI | |
| FCR pulling sensation | Trapeziectomy with LRTI | |
| Temporary parasthesia | Simple trapeziectomy, trapeziectomy with LRTI | |
| Transient inflammatory reaction | Trapeziectomy with spacer | |
| Moderate | Reflex sympathetic dystrophy | Simple trapeziectomy, trapeziectomy with LRTI, |
| Trigger thumb | Trapeziectomy with LRTI | |
| Tendinitis treated with steroids | Simple trapeziectomy, trapeziectomy with LRTI | |
| De Quervain disease | Simple trapeziectomy, trapeziectomy with LRTI | |
| Neuroma | Simple trapeziectomy, trapeziectomy with LRTI | |
| Radial nerve irritation | Simple trapeziectomy, trapeziectomy with LRTI, | |
| Dynamic radiographic subluxation | Swanson intramedullary implant | |
| Delayed union | Arthrodesis (plate and screw) | |
| Severe | Severe osteoporosis | Artelon spacer |
| Osteonecrosis of trapezium | Artelon spacer | |
| Dislocation | Intramedullary implant | |
| FCR rupture | Trapeziectomy with LRTI | |
| Revision surgery | Trapeziectomy, joint replacement (uncemented and cemented implant), trapeziectomy with button | |
| Osteolysis of stem and cup | Joint replacement (uncemented) |
FCR: flexor carpi radialis; LRTI, ligament reconstruction and tendon interposition.
Complication severity was decided following discussions between both authors. Complications were classified as ‘mild’ if they did not pose a significant risk to patient satisfaction, well being, and general health, and were transient or reversible. Severe complications were those that required medical intervention, were irreversible, or presented a substantial risk to patient well being. Moderate complications were those that were neither mild nor severe.
The results of the Bayesian network meta-analysis demonstrated that simple trapeziectomy with anterior approach was associated with the lowest postoperative incidence of complications (SUCRA value ∼95%). The high SUCRA value indicates a high level of certainty in this ranking. Trapeziectomy with anterior approach was followed by simple trapeziectomy with posterior approach (SUCRA value ∼80%), and ligament reconstruction with palmaris longus and ligament reconstruction with flexor carpi radialis (Weilby technique) ranked equal third (SUCRA value ∼73%). Arthrodesis with plate and screw ranked last (SUCRA value ∼10%) (Fig. 3), indicating that this technique was associated with the greatest incidence of postoperative complications.
Figure 3.
Surface under the cumulative ranking curve (SUCRA) diagram∗ for postoperative complications. APL, abductor pollicus longus; FCR: flexor carpi radialis; LRTI, ligament reconstruction and tendon interposition; PL, palmaris longus; PLDLA, poly-d-lactic acid.
∗Diagram illustrating the ranking and cumulative probability of rankings associated with each surgical technique. Simple trapeziectomy is ranked first, and arthrodesis with plate and screw is ranked last.
Compared to trapeziectomy and trapeziectomy with soft tissue procedure, alternative surgical techniques were associated with the lowest incidence of mild and moderate complications during the immediate (0–3 months) follow-up period. Conversely, alternative surgical techniques were associated with the greatest incidence of moderate and severe complications during the intermediate (3–12 months) follow-up period, and alternative surgical techniques were associated with the greatest incidence of severe complications after 12 months follow-up. Collectively, these results indicating a more severe distribution of complications after alternative surgical techniques compared to simple trapeziectomy or trapeziectomy with soft tissue procedure (Supplemental Fig. S2, available on the Journal’s website at www.jhsgo.org).
Results from the Kruskal-Wallis analysis demonstrated that alternative surgical techniques (joint replacement, spacer interposition, arthrodesis, and trapeziectomy with button) were associated with a greater severity of complications compared to simple trapeziectomy and trapeziectomy with soft tissue procedure during the 3- to 12-month follow-up period (Table 3).
Table 3.
Analysis of Complication Severity Across All Time Periods
| Complication Severity | Kruskal-Wallis H† | P Value | ||
|---|---|---|---|---|
| 0–3 mo | Trapeziectomy | 10% mild, 22% moderate, and 0.9% severe | 22.92 | <.001∗ |
| LRTI | 21% mild, 17% moderate, and 4% severe | |||
| Alternative surgical technique | 2% mild, 1% moderate, and 4% severe | |||
| 3–12 mo | Trapeziectomy | 5% mild, 9% moderate, and 3% severe | 46.82 | <.001∗ |
| LRTI | 9% mild, 9% moderate, and 4% severe | |||
| Alternative surgical technique | 7% mild, 18% moderate, and 14% severe | |||
| 12+ mo | Trapeziectomy | 10% mild, 1% moderate, and 3% severe | 7.01 | .03∗ |
| LRTI | 4% mild, 4% moderate, and 4% severe | |||
| Alternative surgical technique | 2% mild, 2% moderate, and 4% severe |
LRTI, ligament reconstruction and tendon interposition.
Indicates statistically significant result (p≤0.05).
The Kruskal-Wallis H statistic is calculated for each comparison of complication severity between surgical techniques (0–3 months, 3–12 months, 12+ months). The H statistic indicates the degree of dissimilarity between complication severity. A H statistic >5.99 indicates that the severity of complications is affected by choice of surgical procedure (trapeziectomy, LRTI, or alternative surgical technique).
A total of eight studies were included in the analysis of postoperative functional improvement. Uncemented joint replacement was associated with the greatest improvement (decrease) in DASH score (mean difference: –36.41; SD: 5.20), followed by simple trapeziectomy (mean difference: –26, SD: 4.47) (Table 4). Arthrodesis with plate and screw (a surgical technique that fuses the joint, eliminating range of motion across the thumb carpometacarpal joint) demonstrated no postoperative functional benefit; patients undergoing this technique had the same mean DASH score pre- and after surgery (mean difference: 0; SD: 12.84).
Table 4.
DASH Scores After CMC OA Surgery
| Technique | No. of Studies | Total No. of Hands | Mean Difference in DASH Score | SD |
|---|---|---|---|---|
| Uncemented joint replacement | 1 | 20 | –36.41 | 5.20 |
| Trapeziectomy | 2 | 99 | –26 | 4.47 |
| Trapeziectomy with button | 1 | 37 | –30.5∗ | 33.3 |
| LRTI (FCR) – Burton-Pellegrini | 6 | 228 | –24.27 | 4.35 |
| LRTI (FCR) – Weilby | 4 | 122 | –19.89 | 4.80 |
| Trapeziectomy with allograft interposition | 1 | 29 | –18 | 26.04 |
| Arthrodesis – Plate and screw | 1 | 17 | 0 | 12.84 |
DASH, Disabilities of Arm, Shoulder and Hand questionnaire; FCR, flexor carpi radialis; LRTI, ligament reconstruction and tendon interposition.
Indicates disagreement between average mean difference across included studies and Bayesian network meta-analysis ranking.
Except for trapeziectomy with button, the results of the Bayesian network meta-analysis supported the results obtained from individual RCTs (Fig. 4). Uncemented joint replacement was ranked first in the Bayesian network analysis (SUCRA value ∼90%), followed by simple trapeziectomy (∼70%). Across all surgical techniques, the wide range of SUCRA values (7-90%) indicates a high level of certainty in the Bayesian ranking.
Figure 4.
Surface under the cumulative ranking curve (SUCRA) diagram∗ for function (DASH).
DASH, Disabilities of Arm, Shoulder and Hand questionnaire; FCR, flexor carpi radialis; LRTI, ligament reconstruction and tendon interposition.
∗Diagram illustrating the ranking and cumulative probability of rankings associated with each surgical technique. Uncemented joint replacement is ranked first, and arthrodesis with plate and screw is ranked last.
Discussion
Our results demonstrate that simple trapeziectomy with posterior approach is associated with favorable outcomes regarding pain and function. Uncemented joint replacement ranked first in the analysis of functional improvement; however, this technique was associated with a high complication rate. Compared to simple trapeziectomy and trapeziectomy with concomitant soft tissue procedure, alternative surgical techniques (joint replacement, spacer interposition, arthrodesis, and trapeziectomy with button) were associated with a significantly greater severity of complications between 3 and 12 months. Collectively, our review indicates that there is a lack of strong evidence that supports the superiority of LRTI or other alternative surgical techniques for surgical treatment of CMC OA.
Regarding pain, we found that simple trapeziectomy ranked first with a relatively high SUCRA value of approximately 80% and a mean improvement in VAS score of 6.86. Trapeziectomy is a well-established procedure that eliminates the source of pain in patients with CMC OA by completely resecting the trapezium. Conversely, trapeziectomy with suture-button suspensionplasty ranked last in the analysis of pain, with a SUCRA value of ∼20%. This may be attributed to the biomechanical implications of the suspensionplasty on the surrounding metacarpal and carpal articulations. Excessive tightening of the suspension in this technique has been described as causing pain between the base of the first two metacarpals, which can lead to erosive arthritis in the long term.22 Notably, both trapeziectomy with LRTI (APL) and the Swanson silicone implant ranked similarly in the analysis of pain, with SUCRA values of 68% and 73%, respectively. Both techniques essentially function to preserve the scaphometacarpal space and prevent metacarpal subsidence. Although metacarpal subsidence has been implicated in the development of postoperative pain in patients undergoing trapeziectomy without concomitant soft tissue procedure, our results suggest that prevention of subsidence is not associated with greater benefit in VAS scores.23 This is supported by a 2021 retrospective review which ultimately concluding that the degree of subsidence did not prognosticate pain.24 Reduction in pain is an important marker of surgical success, as intractable pain is the primary indication for surgical treatment of CMC OA. Therefore, given the observed benefit in pain demonstrated by studies investigating simple trapeziectomy in our review, overall, our results support the use of simple trapeziectomy for the surgical management of CMC OA.
Postoperative complications are another important consideration in patients requiring surgical management of CMC OA. We found that trapeziectomy alone (either anterior or posterior approach) was associated with the lowest number of postoperative complications, and arthrodesis with plate and screw was ranked least favorably in the analysis of complications. We also found that alternative surgical techniques were associated with a greater severity of complications compared to both simple trapeziectomy and trapeziectomy with soft tissue procedure between 3 and 12 months, with 14% of patients experiencing a severe complication (including osteonecrosis of the trapezium, cup dislocation, osteolysis of implant, and revision surgery after implant failure) after alternative surgical technique, compared to 3% after simple trapeziectomy and 4% after trapeziectomy with LRTI. Furthermore, regarding carpometacarpal joint replacement specifically, studies investigating uncemented joint replacement included in our review reported up to 30% rate of severe complications over all follow-up periods.25 Our results are corroborated by current literature, which also describes a high rate of relatively severe complications after alternative CMC OA operative techniques. The use of silicone implants has been implicated in a large number of postoperative complications including silicone synovitis and dislocation.26 In reference to CMC arthroplasty, a 2-year longitudinal cohort study observed a 47% implant failure rate, prompting the authors to state that they do not recommend the use of the Elektra.27 This was also supported by a 2020 study observing a 46% revision rate after Elektra implant, with almost all revisions containing signs of metallosis.28 These findings highlight the issue of cup loosening because of metal-on-metal bearing within the implant, as well as the importance of prioritizing pain relief over marginal functional benefits in patients undergoing surgical treatment of CMC OA. There is a lack of evidence describing robust long-term follow-up of joint replacement techniques for CMC OA in current literature.29 Therefore, considering the severe morbidity associated with events including reoperation and implant intolerance, the associated between severe complications and CMC joint replacement techniques observed in our study warrants further research.
With respect to functional improvement, we found that uncemented joint replacement ranked first, with a SUCRA value of ∼90%. This is likely because of the use of DASH scores as a surrogate measure of functional ability, which is strongly influenced by thumb range of motion.30 Unlike other surgical techniques for CMC OA, joint arthroplasty explicitly preserves or improves thumb range of motion and stability by replacing the native biconcave-convex saddle joint with an unconstrained ball and socket prosthesis.31 The increased range of motion supplied by arthroplasty may assist with tasks including turning a key, opening a jar, and changing a lightbulb, all of which are assessed on the DASH questionnaire. Nevertheless, it is important to note that the success of surgical management of CMC OA remains ultimately contingent upon the observed reduction in postoperative pain, and notably, neither of the two studies investigating joint replacement could be included in the analysis of pain. Hansen et al32 only provided pre-operative VAS scores, and Thorkildsen et al25 did not include pain as an outcome at all. The reason for this is unclear; however, this provides the imperative for further research investigating pain as an outcome after CMC joint replacement.
The primary limitation of our study relates to heterogeneity between studies. In addition to methodological variation between operating techniques, postoperative therapy and rehabilitation regimes differ between institutions. However, the hierarchical modeling employed by the Bayesian network meta-analysis explicitly accounts for uncertainty between and within studies, and the reported SUCRA values are calculated while incorporating this uncertainty. Furthermore, the average follow-up period of RCTs in our study was 2 years and 5 months. Our results represent short-term outcomes after surgical treatment of CMC OA and may underrepresent the morbidity of long-term complications including hardware, including implant failure, metallosis, and silicone disintegration. Finally, we were unable to assess all surgical techniques, including denervation, arthroplasty and arthroscopic management of CMC OA. Denervation for CMC OA has demonstrated promising results in case series, and preliminary results from both arthroplasty and arthroscopic management of CMC OA have demonstrated pain relief while obviating extensive surgical trauma.33,34 However, given the paucity of literature investigating these surgical techniques and lack of randomized controlled trials meeting inclusion criteria at the time of the literature search, these newer techniques for CMC OA could not be included in our study.
Ultimately, although the association between increased complexity of surgical technique and increased prevalence of complications has been previously acknowledged in the literature dating back to the 2005 Cochrane review by Wajon et al,35 our review functions to quantify the severity of complications as provide a more granular analysis of complications after a wider range of surgical techniques. Furthermore, with regard to the other primary outcomes of pain and postoperative improvement in hand function, our review further highlights the utility of trapeziectomy alone, demonstrating favorable rankings in both domains. The novelty of our review lies in the inclusion of newer techniques such as uncemented joint replacement and a variety of ligament reconstruction techniques, as well as the use of the updated Bayesian statistical method for data analysis. Overall, our research provides the foundation for further targeted research investigating the concepts of trapeziometacarpal biomechanics, metacarpal subsidence, and implant dynamics in patients undergoing surgical treatment of CMC OA.
Conflicts of Interest
No benefits in any form have been received or will be received related directly to this article.
Acknowledgments
The authors thank Emeritus Professor John McGeachie and Dr. Nazim Khan for their contributions to the final version of this manuscript, and acknowledge the support provided by The University of Western Australia during the completion of this project.
Supplementary Data
Risk of bias assessment for included studies.
∗ Summary of risk of bias assessment in the domains of detection, selection, performance, attrition, and reporting bias. Key: [+] = low risk of bias detected; [?] = indeterminate risk of bias detected; [-] = high risk of bias detected.
Trends in complication severity over time.
∗ Prevalence of complications [y axis] according to time period (0-3, 3-12, 12+ months) [x-axis], separated into mild, moderate, and severe complications.
Characteristics of included studies.
∗ Ligament reconstruction and tendon interposition (LRTI), Burton-Pellegrini technique (BP), Flexor carpi radialis tendon (FCR), Abductor pollicis longus tendon (APL), Poly-L/D-lactic acid (PLDLA).
Outcomes measured in each study.
1IQR: interquartile range; SD: standard deviation; SE: Standard error
2VAS: Visual analogue scale; PRWHE: Patient-reported wrist and hand outcome evaluation; DASH: Disabilities of arm, shoulder, and hand questionnaire; qDASH: quickDASH, SF-12: 12-item short form survey; MHQ: Michigan hand outcomes questionnaire; PEM: Patient evaluation measure∗: included in analyses.
References
- 1.Teunissen J.S., Wouters R.M., Bierma-Zeintra S.M.A., et al. The prevalence, incidence, and progression of radiographic thumb base osteoarthritis in a population-based cohort: the Rotterdam Study. Osteoarthritis Cartilage. 2022;30(4):578–585. doi: 10.1016/j.joca.2022.01.003. [DOI] [PubMed] [Google Scholar]
- 2.O’Brien V., Johnson J., Pisano K., Enke A. Dynamic stabilization of the painful thumb: A historical and evidence-informed synthesis. J Hand Ther. 2022;35(3):388–399. doi: 10.1016/j.jht.2022.06.007. [DOI] [PubMed] [Google Scholar]
- 3.Wouters R.M., Vranceanu A.M., Slijper H.P., et al. Patients with thumb-base osteoarthritis scheduled for surgery have more symptoms, worse psychological profile, and higher expectations than nonsurgical counterparts: a large cohort analysis. Clin Orthop Relat Res. 2019;477(12):2735–2746. doi: 10.1097/CORR.0000000000000897. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gillis J., Calder K., Williams J. Review of thumb carpometacarpal arthritis classification, treatment and outcomes. Can J Plast Surg. 2011;19(4):134–138. doi: 10.1177/229255031101900409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kloppenburg M., Kroon F.P., Blanco F.J., et al. 2018 update of the EULAR recommendations for the management of hand osteoarthritis. Ann Rheum Dis. 2019;78(1):16–24. doi: 10.1136/annrheumdis-2018-213826. [DOI] [PubMed] [Google Scholar]
- 6.Lane J.C., Craig R., Rees J.L., et al. Low rates of serious complications and further procedures following surgery for base of thumb osteoarthritis: analysis of a national cohort of 43 076 surgeries. BMJ Open. 2021;11(7) doi: 10.1136/bmjopen-2020-045614. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Yoon A.P., Hutton D.W., Chung K.C. Cost-effectiveness of surgical treatment of thumb carpometacarpal joint arthritis: a value of information study. Cost Eff Resour Alloc. 2023;21(1):28. doi: 10.1186/s12962-023-00438-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Liu Q., Xu B., Lyu H., Lee J.H. Differences between simple trapeziectomy and trapeziectomy with ligament reconstruction and tendon interposition for the treatment of trapeziometacarpal osteoarthritis: a systematic review and meta-analysis. Arch Orthop Trauma Surg. 2022;142(6):987–996. doi: 10.1007/s00402-020-03707-w. [DOI] [PubMed] [Google Scholar]
- 9.Molin U., Evans K., Wilcke M. Does proximal migration of the first metacarpal correlate with remaining pain after trapeziectomy? J Plast Surg Hand Surg. 2022;56(1):11–15. doi: 10.1080/2000656X.2021.1898975. [DOI] [PubMed] [Google Scholar]
- 10.Wolfe S.W., Pederson W.C., Kozin S.H., Cohen M.S. Green's operative hand surgery. Eighth edition. Operative Hand Surgery. Elsevier; 2022. [Google Scholar]
- 11.Li Y.K., White C., Ignacy T.A., Thoma A. Comparison of trapeziectomy and trapeziectomy with ligament reconstruction and tendon interposition: a systematic literature review. Plast Reconstr Surg. 2011;128(1):199–207. doi: 10.1097/PRS.0b013e318217435a. [DOI] [PubMed] [Google Scholar]
- 12.Qureshi M.K., Halim U.A., Khaled A.S., Roche S.J., Arshad M.S. Trapeziectomy with ligament reconstruction and tendon interposition versus trapeziometacarpal joint replacement for thumb carpometacarpal osteoarthritis: a systematic review and meta-analysis. J Wrist Surg. 2022;11(3):272–278. doi: 10.1055/s-0041-1731818. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Higgins J.P.T., Altman D.G., Gøtzsche P.C., et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ. 2011;343 doi: 10.1136/bmj.d5928. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Luo D., Wan X., Liu J., Tong T. Optimally estimating the sample mean from the sample size, median, mid-range, and/or mid-quartile range. Stat Methods Med Res. 2018;27(6):1785–1805. doi: 10.1177/0962280216669183. [DOI] [PubMed] [Google Scholar]
- 15.Wan X., Wang W., Liu J., Tong T. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol. 2014;14(1):135. doi: 10.1186/1471-2288-14-135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Review Manager (RevMan) 2020. https://training.cochrane.org/online-learning/core-software/revman/revman-5-download
- 17.Owen R.K., Bradbury N., Xin Y., Cooper N., Sutton A. MetaInsight: an interactive web-based tool for analyzing, interrogating, and visualizing network meta-analyses using R-shiny and netmeta. Res Synth Methods. 2019;10(4):569–581. doi: 10.1002/jrsm.1373. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Nevill C.R., Cooper N.J., Sutton A.J. A multifaceted graphical display, including treatment ranking, was developed to aid interpretation of network meta-analysis. J Clin Epidemiol. 2023;157:83–91. doi: 10.1016/j.jclinepi.2023.02.016. [DOI] [PubMed] [Google Scholar]
- 19.Rücker G., Schwarzer G. Ranking treatments in frequentist network meta-analysis works without resampling methods. BMC Med Res Methodol. 2015;15(1):58. doi: 10.1186/s12874-015-0060-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.IBM SPSS Statistics. https://www.ibm.com/products/spss-statistics
- 21.Burton R., Evarts C. Churchill Livingstone; 1983. Surgery of the Musculoskeletal System. [Google Scholar]
- 22.Pistorio A.L., Moore J.B. Lessons learned: trapeziectomy and suture suspension arthroplasty for thumb carpometacarpal osteoarthritis. J Hand Microsurg. 2022;14(3):233–239. doi: 10.1055/s-0040-1716607. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Miller A.J., Jones C.M., Martin D.P., et al. Reliability of metacarpal subsidence measurements after thumb carpometacarpal joint arthroplasty. J Hand Microsurg. 2018;10(1):22–25. doi: 10.1055/s-0037-1618912. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Meyers A., Krebs J., Rampazzo A., Gharb B.B. Evolution of metacarpal subsidence following trapeziectomy. Plastic and Reconstructive Surgery – Global Open. 2021;9(10S):96. doi: 10.1097/PRS.0000000000009921. [DOI] [PubMed] [Google Scholar]
- 25.Thorkildsen R.D., Røkkum M. Trapeziectomy with LRTI or joint replacement for CMC1 arthritis, a randomised controlled trial. J Plast Surg Hand Surg. 2019;53(6):361–369. doi: 10.1080/2000656X.2019.1635490. [DOI] [PubMed] [Google Scholar]
- 26.Minami A., Iwasaki N., Kutsumi K., Suenaga N., Yasuda K. A long-term follow-up of silicone-rubber interposition arthroplasty for osteoarthritis of the thumb carpometacarpal joint. Hand Surgery. 2005;10(01):77–82. doi: 10.1142/S0218810405002607. [DOI] [PubMed] [Google Scholar]
- 27.Hernández-Cortés P., Pajares-López M., Robles-Molina M.J., Gómez-Sánchez R., Toledo-Romero M.A., De Torres-Urrea J. Two-year outcomes of Elektra prosthesis for trapeziometacarpal osteoarthritis: a longitudinal cohort study. J Hand Surg Eur Vol. 2012;37(2):130–137. doi: 10.1177/1753193411414505. [DOI] [PubMed] [Google Scholar]
- 28.Froschauer S.M., Holzbauer M., Hager D., Schnelzer R., Kwasny O., Duscher D. Elektra prosthesis versus resection-suspension arthroplasty for thumb carpometacarpal osteoarthritis: a long-term cohort study. J Hand Surg Eur Vol. 2020;45(5):452–457. doi: 10.1177/1753193419873230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Duerinckx J., Verstreken F. Total joint replacement for osteoarthritis of the carpometacarpal joint of the thumb: why and how? EFORT Open Rev. 2022;7(6):349–355. doi: 10.1530/EOR-22-0027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Yang Z., Lim P.P.H., Teo S.H., Chen H., Qiu H., Pua Y.H. Association of wrist and forearm range of motion measures with self-reported functional scores amongst patients with distal radius fractures: a longitudinal study. BMC Musculoskelet Disord. 2018;19(1):142. doi: 10.1186/s12891-018-2065-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Holme T.J., Karbowiak M., Clements J., Sharma R., Craik J., Ellahee N. Thumb CMCJ prosthetic total joint replacement: a systematic review. EFORT Open Rev. 2021;6(5):316–330. doi: 10.1302/2058-5241.6.200152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Hansen T.B., Stilling M. Equally good fixation of cemented and uncemented cups in total trapeziometacarpal joint prostheses. A randomized clinical RSA study with 2-year follow-up. Acta orthopaedica. 2013;84(1):98–105. doi: 10.3109/17453674.2013.765625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Donato D., Abunimer A.M., Abou-Al-Shaar H., Willcockson J., Frazer L., Mahan M.A. First carpometacarpal joint denervation for primary osteoarthritis: technique and outcomes. World Neurosurgery. 2019;122:e1374–e1380. doi: 10.1016/j.wneu.2018.11.061. [DOI] [PubMed] [Google Scholar]
- 34.Wong C.W.-Y., Ho P.C. Arthroscopic management of thumb carpometacarpal joint arthritis. Hand Clin. 2017;33(4):795–812. doi: 10.1016/j.hcl.2017.07.007. [DOI] [PubMed] [Google Scholar]
- 35.Wajon A., Ada L., Edmunds I. Surgery for thumb (trapeziometacarpal joint) osteoarthritis. Cochrane Database Syst Rev. 2005;(4) doi: 10.1002/14651858.CD004631.pub2. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Risk of bias assessment for included studies.
∗ Summary of risk of bias assessment in the domains of detection, selection, performance, attrition, and reporting bias. Key: [+] = low risk of bias detected; [?] = indeterminate risk of bias detected; [-] = high risk of bias detected.
Trends in complication severity over time.
∗ Prevalence of complications [y axis] according to time period (0-3, 3-12, 12+ months) [x-axis], separated into mild, moderate, and severe complications.
Characteristics of included studies.
∗ Ligament reconstruction and tendon interposition (LRTI), Burton-Pellegrini technique (BP), Flexor carpi radialis tendon (FCR), Abductor pollicis longus tendon (APL), Poly-L/D-lactic acid (PLDLA).
Outcomes measured in each study.
1IQR: interquartile range; SD: standard deviation; SE: Standard error
2VAS: Visual analogue scale; PRWHE: Patient-reported wrist and hand outcome evaluation; DASH: Disabilities of arm, shoulder, and hand questionnaire; qDASH: quickDASH, SF-12: 12-item short form survey; MHQ: Michigan hand outcomes questionnaire; PEM: Patient evaluation measure∗: included in analyses.