Total wrist arthroplasty has evolved over several decades and has made gains in outcomes and popularity. In contrast, wrist hemiarthroplasty is a more recent innovation. A systematic review of available literature has been performed to assess the clinical effectiveness of both these treatment options. A total of 38 studies were included in the qualitative analysis, representing 1,543 total wrist arthroplasty and 117 hemiarthroplasty index operations. The effects of surgery on pain, function, range of motion, and grip strength were examined. Rates of complications and implant survival were also assessed. The introduction of third- and fourth-generation implant total arthroplasty designs has led to improvements in outcomes and implant survival, although reporting of outcomes was not consistent across all studies. Only early- to mid-term outcome reporting was available for hemiarthroplasty, and results were more variable. Ongoing performance monitoring remains vital for both procedures.
Key wordswrist arthritis – arthroplasty – hemiarthroplasty – outcome – systematic review
24 Systematic Review of Wrist Arthroplasty
24.1.1 Evolution of Total Wrist Arthroplasty
Wrist prostheses have undergone significant evolution since the earliest implant designs. The first generation of implant to receive wide commercial use was the Swanson silicone prosthesis, a flexible stemmed hinged spacer introduced in 1967. 1 Early results were promising with good pain relief and functional range of motion, but longer follow-up revealed high implant failure/fracture rates of up to 65%, and a soft tissue reactive silicone synovitis in approximately 30% of cases. 2 , 3 , 4
Several other prostheses followed, including the Meuli wrist arthroplasty system in 1972, 5 the Volz total wrist arthroplasty (TWA) in 1973, 6 , 7 the Trispherical TWA in 1977, 8 the Biaxial TWA in 1989, 9 the Destot prosthesis in 1991, 10 and the Anatomic Physiologic Wrist in 1996. 11 There were three iterations of the Meuli wrist prosthesis (I–III). The first was a cemented ball-and-socket metal-on-metal articulation. 12 The ball was later changed to create a metal-on-polyethylene bearing surface and the center of rotation was altered to improve wrist alignment and stability (Meuli II). 5 Due to persisting high rates of dislocation and loosening, Meuli III was developed in 1986, with a cementless fixation and a bearing surface altered to create a metal ball on an ultra-high molecular weight polyethylene (UHMWPE) socket. 13 , 14
Implants such as the Meuli II and Volz were considered the second generation of TWA, with the defining features being radial and carpal components with a ball-and-socket or hemispherical articulation. 15 , 16 , 17 The third generation of implants, such as the Meuli III, Universal, and Biaxial arthroplasties, attempted to better reproduce the center of wrist rotation so as to improve stability and reduce dislocation. 13 , 14 , 18 , 19 The Biaxial implant consisted of a hemispherical metal-on-polyethylene-bearing design and was developed at the Mayo clinic in the late 1970s. Results were first published by Cobb and Beckenbaugh in 1996. 9 The Universal wrist was developed by Menon; the results of the first series were published in 1998. 18 The Universal consisted of titanium radial and carpal components with a toroidal-shaped high-density polyethylene-bearing surface. The carpal component was stabilized with three screws into the carpus.
24.1.2 Current Implants
The Motec (Fig. 24.1; Swemac Innovation AB), Universal 2 (Fig. 24.2; Integra LifeSciences), Freedom (Fig. 24.3; Integra LifeSciences), Re-Motion (Stryker), and Maestro (Biomet) are examples of fourth-generation implants. These newer implants typically require less bone resection, with removal of only the proximal carpal row. Carpal fixation distally is mostly achieved with bone screws, and proximally with a press-fit into the radius. The implants are typically porous coated to enable osseointegration. 20 , 21
The Motec prosthesis (Fig. 24.1), although considered a fourth-generation implant, has some design features reflecting earlier generation principles with an articulation which is a metal-on-metal ball-and-socket bearing. 16 , 22 A new screw fixation concept was introduced for the radial and carpal components. The screws are manufactured from a grit-blasted titanium alloy, coated with a resorbable calcium phosphate (Bonit; DOT Medical Solutions Laboratories Gmbh, Rostock, Germany) to encourage osseointegration. 22 The Motec, previously called the Gibbon, evolved from the Elos prosthesis, which was in use between 2000 and 2005. The performance of the Elos and Gibbon has been described by Krukhaug et al in their report on the Norwegian Arthroplasty Register. 23
The Universal was initially designed with a toroidal articulation but was later changed to an ellipsoidal shape to improve stability and reduce the risk of dislocation. 24 This was one of the changes that led to the development of the contemporary Universal 2 (Fig. 24.2), which features a titanium alloy carpal component with two screws and a central peg, and a cobalt-chrome alloy radial component. The Universal 2 wrist arthroplasty has been further modified to produce the Freedom wrist arthroplasty (Fig. 24.3). The design changes include a shorter, more tapered carpal stem, a lower profile radial tray with a 5-degree rotation relative to the radial stem for slight supination of the wrist at rest and a shorter less invasive radial stem. The Re-Motion arthroplasty system (see Fig. 25.1) has cobalt-chrome alloy radial and carpal components, with titanium coated to the undersurface of the carpal component. The articulation is a metal on UHMWPE. The results of the first patient series was published by Herzberg et al. 25 The Maestro wrist system incorporates an uncemented titanium alloy stem articulating with a cobalt-chrome radial body. An UHMWPE-bearing surface is fixed to the radial body. The carpal component is a monoblock of cobalt-chrome with a titanium alloy central-capitate peg.
Currently, the four commercially available wrist implants in the United States are the Universal 2, Freedom, Re-Motion, and Maestro. In Europe the Motec is also approved for use. More recently, a new wrist implant was introduced, named the Prosthelast (Argomedical). 26 It consists of a titanium radial implant fixed to an elastic intramedullary wire that rests on the subchondral bone of the radial head, aiming to preventing axial migration of the radial component. The carpal component has a polyethylene-bearing surface which attaches to a titanium plate. Only preliminary data with short-term follow-up are available, so this implant was not included in this study.
Wrist arthroplasties have been analyzed in several reviews. 27 , 28 , 29 , 30 Cavaliere and Chung performed a comparative review of wrist arthroplasty versus wrist arthrodesis. 27 No pain or mild pain was experienced by 90% of patients in the arthroplasty group, but the rates of major complication were higher in the arthroplasty group (25%) compared to the arthrodesis group (13%). Newer third-generation prostheses had lower major complication rates (21%) than older generation implants. The systematic review by Yeoh and Tourret had similar findings with arthroplasties having higher complication rates. 29 A more recent systematic review by Berber et al demonstrated that newer fourth-generation implants had lower complication rates (range 0.1–2.9%) than earlier designs (range 0.2–8.1%; p=0.002). 30 Maintenance of a range of motion is seen as the clear advantage of wrist arthroplasty over arthrodesis, yet this benefit is not always shown in objective assessment. A functional range of motion, as defined by Palmer et al is 5 degrees of flexion, 30 degrees of extension, 10 degrees of radial deviation, and 15 degrees of ulnar deviation, 31 was achieved in only 3 of the 14 studies in the review by Cavaliere and Chung with similar findings reported by Berber et al. 27 , 30
Wrist hemiarthroplasty is an emerging management option for wrist arthritis and wrist trauma. To date there have been several publications reporting early to mid-term outcomes. 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 Studies have also reported on the early clinical results of hemiarthroplasty in elderly patients with acute non-reconstructable distal radial fractures, as first described by Roux. 32 , 36 , 37 , 38 , 39 , 40 Anneberg et al reported on the outcomes of a new hemiarthroplasty system, KinematX (Extremity Medical, LLC). 35 The remaining studies typically utilized the radial or carpal components of a current TWA system.
Since the most recent review by Berber et al, 30 there have been several new publications with greater data available on current fourth-generation wrist arthroplasty systems, as well as on wrist hemiarthroplasties.
The proposed research question forming the basis for this review is: “What is the clinical effectiveness of TWA and wrist hemiarthroplasty?”
The objective of this review was to evaluate the clinical effectiveness of TWA, defined as the effect of TWA on the specific outcomes listed below:
Outcome 1: Pain.
Outcome 2: Secondary measures of function (e.g., grip strength) and quality of life.
Outcome 3: The frequency of treatment failure/implant survival and adverse events.
We also wished to assess whether the use of fourth-generation wrist arthroplasty systems was associated with improved clinical effectiveness.
A study protocol was created prior to undertaking this review. This was modified from a previous protocol published on the National Institute for Health Research PROSPERO Database (CRD42017067377). Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were adhered to.
24.3.1 Study Inclusion Criteria
Types of Participants
Studies were included if participants were adults over the age of 16 years and of any gender. The underlying clinical diagnosis was arthritis of the wrist. The cause of arthritis was kept broad to reflect common daily practice in most orthopaedic clinics, including both inflammatory and noninflammatory arthritis.
Types of Interventions
The two interventions were TWA and wrist hemiarthroplasty. There have been four generations of TWA, with only the latest generation of implants still commercially available. Subgroup analysis by generation was performed to assess whether the performance of the latest implants has improved.
Types of Outcome Measures
The core measures used as a basis for defining the primary and secondary outcomes in this study originated from recommendations established by the OMERACT (Outcome Measures in Rheumatoid Arthritis Clinical Trials) conference in 1992. These were first approved by the American College of Rheumatology 41 and later published under WHO and ILAR endorsement. 42 The core end points were also refined for osteoarthritis in general 43 and hand osteoarthritis 44 in particular.
Primary and Secondary Outcome Measures
Pain relief was chosen as the primary outcome in this study. Pain is commonly reported as a continuous variable on a “visual analog scale” (VAS) or on a categorical scale. Secondary outcome measures included function, grip strength, wrist motion, adverse events, and implant survival.
Inclusion and Exclusion Criteria
The studies chosen for inclusion in this review include randomized control trials, cohort studies, case–control studies, and case series. Case reports and abstracts were excluded from this review. Studies with fewer than five patients and studies failing to report on any of the primary or secondary outcomes were excluded. The minimum duration of follow-up was chosen as 1 year.
24.3.2 Search Methods
Several databases were searched on the 20th of May 2019, including OVID Medline (dates: 1946 to present), OVID Excerpta Medica Database (EMBASE) (dates: 1974 to present), Cochrane Central Register of Controlled Trials (Issue 6 of 12, June 2019), NICE Database, Cumulative Index to Nursing and Allied Health Literature (CINAHL) (1981 to present), and British Nursing Index (BNI) (1992 to present). Trial registers were also searched for relevant studies including ClinicalTrials.gov and the World Health Organization International Clinical Trials Registry Platform (ICTRP).
The reference lists from the papers identified in the above searches were reviewed for potential studies of interest. The “PICOS” elements (Population, Intervention, Comparator, and Outcomes) were used to construct an effective search strategy. Other search filters included studies published in the English language only; animal or cadaveric studies were excluded. No date restrictions were applied to maximize search numbers.
24.3.3 Data Collection and Analysis
Study Selection, Data Extraction, and Management
Studies were selected through two stages of screening. The first stage involved a review of the study title and abstracts to remove obviously irrelevant articles. The full text of the remaining articles was then reviewed in more detail by the study authors. The basic dataset for extraction was adapted from recommendations in the Cochrane Handbook for Systematic Review of Interventions. 45
Assessment of Risk of Bias and Quality of Evidence
The GRADE system was chosen to perform a quality appraisal of case series. 46 This system specifies four levels of quality (high, moderate, low, very low) based on five factors including study design (randomized controlled trials vs. observational studies), study quality (bias, loss to follow-up, sparse data), consistency of results (degree of consistency of effect between or within studies), directness of evidence (generalizability of population and outcomes to population of interest), and effect size.
24.4.1 Description of Studies
The results returned from the database searches are summarized in Fig. 24.4. Of the 38 articles, 31 reported the outcomes of wrist arthroplasty, 10 , 11 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 23 , 33 , 47 , 48 , 49 , 50 , 51 , 52 , 53 , 54 , 55 , 56 , 57 , 58 , 59 , 60 , 61 , 62 , 63 , 64 , 65 two undertook a retrospective matched cohort review of arthroplasty against arthrodesis cases 66 , 67 and five reviewed the outcomes of wrist hemiarthroplasty. 32 , 33 , 34 , 35 , 37
Table 24.1 summarizes the overall patient demographics of both the TWA and hemiarthroplasty patient cohorts.
Total Wrist Arthroplasty
The underlying wrist pathology in the TWA studies was a mix of inflammatory and noninflammatory causes in the majority of the studies. However, in some, a TWA was performed only on patients with noninflammatory wrist arthritis. 10 , 16 , 60 , 62 , 65 , 67 Two studies undertook a retrospective matched cohort review of arthroplasty against arthrodesis. 66 , 67 Both were included as the TWA data could be extracted for analysis.
Common reasons for excluding studies from this analysis included patient follow-up of less than 1 year and a lack of outcome measurements and reporting of revision wrist arthroplasty. Four studies were excluded as more recent data have been published. 8 , 56 , 68 , 69 Botero et al from Strasbourg, France was the first to report on a new TWA implant (Prosthelast). 26 Patient follow-up was less than 1 year and so the paper was rejected from this review.
The majority of wrist hemiarthroplasties were performed on patients with noninflammatory wrist pathology (Table 24.1). In two studies, wrist hemiarthroplasty was performed on elderly patients with acute non-reconstructable distal radius fractures. 32 , 38 This was first reported by Roux et al; however, the paper was in French and hence is not included in this review. 39 This was also reported by Vergnenegre et al; this paper was also rejected as patient follow-up was less than 1 year. 40 There was an updated publication by Herzberg et al on earlier results and therefore only this most recent article was included. 36 , 37
Wrist Arthroplasty Implants
Total Wrist Arthroplasty
Table 24.2 provides a summary of the TWA studies. Fourteen different implants were used, ranging from earlier second-generation prostheses such as the Meuli and Volz to contemporary fourth-generation systems such as the Universal 2, Maestro, Re-Motion, and Motec.
Wrist hemiarthroplasty was reported in five studies summarized in Table 24.3. Herzberg et al and Ichihara et al both reported on wrist hemiarthroplasty for acute non-reconstructable distal radial fractures in elderly patients. 32 , 37 The implants included the radial (DRH) or carpal components (CH) of pre-existing TWA systems such as the radial component of the Maestro 33 , 34 or Re-Motion 33 , 37 wrist systems; the carpal component of the Maestro 33 ; and more recently the KinematX 35 , the Cobra 37 (GroupeLepine), and the radial component of the Prosthelast system. 32
Quality of Evidence
Several study authors were involved in product design 10 , 14 , 16 , 18 , 55 , 59 , 65 and Divelbiss et al received benefits for their work. 54 The GRADE quality assessment was of “moderate grade” in several studies as the risk of bias was low and the reported outcome effect sizes were sufficient. 21 , 23 , 46 , 49 , 56 , 57 , 58 , 60 , 62 , 65 , 66 , 67 The remaining studies were “very low” to “low” grade. These studies were generally retrospective observational studies, with no blinding and often with missing data.
24.4.2 Effects of Interventions
Pre- and postoperative pain scores were often not reported. Usually scores were only measured at the latest follow-up. Table 24.4 presents studies, which reported both pre- and postoperative scores. The improvement in pain scores following surgery was significant in several cases (Table 24.4). 17 , 21 , 47 , 48 , 56 , 58 , 59 , 62 , 63 , 65
The study by Radmer et al was a clear outlier from the general trend. 11 Total wrist replacement was performed with the APH prosthesis, an uncemented implant that demonstrated good early results at 18 months. Follow-up at a mean of 52 (range 24–73) months, however, demonstrated catastrophic failure of all cases with a 100% revision to arthrodesis.
Patient-reported pain scores were poorly measured in the hemiarthroplasty studies. In the study by Ichihara et al using the Prosthelast implant immediately following a distal radius fracture, patients had a mean pain score of 2.8 (0–10 severe) at a mean follow-up of 32 (range 24–42) months. 32 Herzberg et al using the radial component of the Re-Motion and the Cobra implant reported a mean postoperative pain score of 1 (0–10 severe) following surgery for acute distal radial fractures. 37 Herzberg et al reported on a further cohort of four patients in whom they performed a hemiarthroplasty for persistent symptoms after a distal radial fracture at a mean of 3 (range 1–6) months post injury. 37 They reported an improvement in pain scores from a mean of 6 to 2.5. Of note, in both these studies, the hemiarthroplasty was performed in elderly but independent patients with apparently unreconstructible distal radial fractures.
Table 24.5 provides a summary of the TWA studies which used the DASH or Quick-DASH, therefore allowing more direct comparison between implants. Several arthroplasty studies reported a significant improvement in outcome scores following surgery (Table 24.5). 54 , 56 , 57 , 59 , 61 , 65 Sagerfors et al. observed this improvement with several different implants up to 5 years after surgery in outcome measures also including the Canadian occupational performance measure (COPM) and patient-reported wrist evaluation (PRWE). 21 Herzberg et al and Boeckstyns et al reported no statistically significant difference in the Quick-DASH score for wrist arthroplasty performed for inflammatory or noninflammatory arthritis. 56 , 64
The hemiarthroplasty studies typically reported DASH or Quick-DASH scores, and in nontrauma cases, pre- and postoperative scores were provided (Table 24.6). Trends were difficult to identify because of the small number of studies to date. Anneberg et al demonstrated an improvement in the DASH score using the KinematX hemiarthroplasty in 20 patients with wrist arthritis at a mean follow-up of 4.1 (range 2.3–5.3) years. 35 Huish et al, however, had contrasting findings 34 : they used the carpal component of the Maestro wrist system in 11 patients with advanced wrist arthritis. They found no significant improvement in the DASH score following surgery. A high failure rate was observed in this study (45%; 5 of 11 cases).