2.2 Shoulder - Arthritis

Search Terms

(shoulder OR glenohumeral) AND (arthritis OR arthroplasty) AND (incidence OR prevalence OR epidemiolog* OR treatment ) AND (outcomes OR pain OR function)

2.2.1 Definition and epidemiology

Osteoarthritis of the glenohumeral joint is characterised by loss of articular cartilage, glenoid bone loss, and increased glenoid retroversion with posterior subluxation of the humeral head in advanced cases (Stephens et al. 2015). Patients can present with: (a) primary osteoarthritis that results from age-related biochemical changes of cartilage leading to structural weakening, or (b) secondary osteoarthritis, in which an antecedent event, such as a large rotator cuff tear or trauma, causes degradation of normal cartilage that is due to abnormal biomechanics (Lin et al. 2016). Inflammation of the surrounding soft tissues is often present and further contributes to the pain caused by the disease process (Menge et al. 2014).

Osteoarthritis causes significant pain, functional limitation and disability. It has an estimated prevalence of between 4% to 26% (Bull et al. 2016), affecting up to one-third of individuals over 60 years of age (Menge et al. 2014). The progression of shoulder osteoarthritis is most commonly described using the Walch classification of glenoid morphology (Vo et al. 2017):

• Type A, centred or symmetric arthritis without posterior subluxation of the humeral head:
  • Type A1 has minor central wear or erosion
  • Type A2 has severe or major central wear or erosion
• Type B, asymmetric arthritis with posterior subluxation of the humeral head:
  • Type B1 has no obvious glenoid erosion with posterior joint space narrowing, subchondral sclerosis, and osteophytes
  • Type B2 has apparent or obvious erosion of the posterior glenoid forming a biconcave appearance of the glenoid
• Type C, glenoid retroversion greater than 25deg (dysplastic in origin) regardless of glenoid erosion or the location of the humeral head with regard to the glenoid.

Glenoid type has been associated with glenohumeral arthritis outcomes.

2.2.2 Treatment overview

Total Shoulder Arthroplasty (anatomic): Surgical implantation of an anatomic glenohumeral joint replacement that addresses pathologic conditions involving both sides of the joint (Lin et al. 2016). Anatomic total shoulder arthroplasty (aTSA) requires an intact rotator cuff because the resultant anatomic joint replacement relies on native soft-tissue structures for mobility and longevity (Lin et al. 2016). It also requires adequate glenoid bone stock to allow anatomic placement of the glenoid prosthetic component so as to prevent eventual glenoid component loosening and prosthetic failure (Lin et al. 2016).

Reverse Total Shoulder Arthroplasty: The reverse total shoulder arthroplasty (rTSA) uses an implant that reverses the ball and socket of the shoulder joint. This procedure is most commonly used in cases of rotator cuff arthropathy, and may be used in select cases where rotator cuff dysfunction is suspected, or where inadequate stability of the anatomic glenoid component is anticipated due to glenoid morphology or bone loss (Chalmers & Keener 2016; Streit et al. 2017). The glenoid components are modular and consist of a base (metaglene) with a coarse surface that is affixed to the native glenoid by locking and nonlocking screws and a ball (glenosphere) that is secured to the base plate (Lin et al. 2016). The humeral component consists of a metal stem that is monoblock or modular and is cemented or noncemented, as well as a customizable polyethylene-lined cup that allows for different degrees of constrained articulation (Lin et al. 2016).

Postoperative rehabilitation: Postoperative rehabilitation is routinely prescribed for patients undergoing surgical procedures for shoulder arthritis. Patients are immobilised in a sling for 6 weeks and instructed to undertake passive range of motion exercises with no external rotation beyond 0 degrees. After 6 weeks, return to activity is introduced as tolerated.

2.2.3 Treatment Outcomes Summary

Outcomes most commonly reported in the literature include revision, SST score, WOOS, ASES, Constant score, and RoM (Bull et al. 2016). Tashjian and colleagues (2016) suggest that a 1.4- point change on the VAS for pain, a 20.9-point change in the ASES score, and a 2.4-point change in the SST score achieve a clinically important improvement after shoulder arthroplasty at 2 years follow-up. Similarly, Werner and colleagues (2016) suggest that a 9-point improvement in their ASES score indicates the minimal clinically important difference, but those who have at least a 23-point improvement in their ASES score experience a substantial clinical benefit at 2-years follow-up.

A number of factors interplay to influence patient outcome following diagnosis with osteoarthritis of the shoulder (Figure 2.2.4), with treatment factors being only one branch. Treatment outcomes are also influenced by patient factors (e.g., history of shoulder issues (Matsen et al. 2016)) and pathology (e.g., preoperative range of motion (Levy et al. 2016)).

In their most recent annual report (2018), the AOANJRR reported 3-year cumulative revision rates of 6% for aTSA with stemmed components, and 3.6% for rTSA for patients with a primary diagnosis of osteoarthritis. Most common reasons for revision of stemmed aTSA were rotator cuff insufficiency, instability/dislocation, and loosening, comprising 23.1%, 22.9 and 16.8% of revisions. Most common reasons for revision of rTSA were instability/dislocation, infection, and loosening, comprising 34.5%, 19.9%, and 18.3% of revisions.

anatomic Total Shoulder Arthroplasty: Shoulder arthroplasty is considered the “gold standard” for surgical management of advanced glenohumeral arthritis in older patients. Use of arthroplasty is more complex in younger (<55 years old) patients due to mixed pathology and higher physical outcome requirements to return to normal function (Johnson et al. 2015). However, the literature suggests that total shoulder arthroscopy provides better clinical outcomes than hemiarthroplasty in the surgical treatment of glenohumeral arthritis in young patients (Sayegh et al. 2015). Overall, total shoulder arthroplasty is the most effective option for symptom relief (Sayegh et al. 2015) and leads to improvement in both shoulder-specific and overall physical wellbeing (Carter et al. 2012).

reverse Total Shoulder Arthroplasty: When performed for cuff tear arthropathy, rTSA tends to result in superior patient-reported outcomes and lower revision rates than aTSA (Samitier et al. 2015). A small number of studies suggest that when performed in select osteoarthritis cases with an intact rotator cuff (for example in the setting of inadequate seating of anatomic TSA glenoid trial components, or in posterior instability), rTSA may result in roughly equivalent outcomes as aTSA (Streit et al. 2017; Steen et al. 2015; Mizuno et al. 2013).

2.2.4 Summary and gap

• The rate of failure and revision following treatment of shoulder arthritis remains relatively high compared to other joint pathology and the factors associated with treatment success or failure remains poorly elucidated
• Evidence identifying the key factors influencing patient outcome following diagnosis and treatment of shoulder arthritis is unclear, particularly in relation to younger populations.

2.2.5 Cohort aims and hypotheses

Primary

In patients presenting with glenohumeral arthritis of any aetiology treated with total shoulder arthroplasty, what are the patient (demographic, anatomical), pathology and management (surgical, rehabilitation) factors associated with treatment success/failure within 2 years of surgery

It is hypothesized that patient, pathology and management factors will be significant factors in identifying patients considered clinical failure as per the a-priori criteria.

Secondary

In patients presenting with glenohumeral arthritis of any aetiology undergoing TSA and standard rehabilitation, with OR without the use of a continuous passive movement device, what is the rate of treatment success or shoulder range of motion within 1 year of surgery?

2.2.6 Cohort inclusion/exclusion criteria

Inclusion Criteria

• Primary diagnosis of Glenohumeral arthritis of any aetiology.
• Glenohumeral arthritis secondary to instability or rotator cuff arthropathy, if arthritis is to be considered the principal pathology to be addressed.
• Undergoing treatment by primary investigator/orthopaedic consultant surgeon.
• Consenting to being part of the Clinical Research Registry

Exclusion Criteria

• Glenohumeral arthritis is present, is deemed secondary to other conditions (except instability or rotator cuff arthropathy, see above).
• Patient-reported outcomes: Clinically diagnosed mental/neurological illness (indicated on GP referral, or identified through secondary referral to psychiatric/neurological specialist) that precludes retrieval of patient feedback via interview, paper or electronic questionnaire (e.g., schizophrenia, dementia, Alzheimer's Disease, personality disorder).
• Revocation of consent for research use of personal data.

2.2.8 Cohort outcome measures

2.2.9 Cohort secondary outcome measures

Active range of motion of the shoulder (glenohumeral + scapulothoracic) Active range of motion of the shoulder (glenohumeral + scapulothoracic) will be assessed with the patient (sitting, standing) with the scapular unrestrained. Testing position will be standardised (recorded at least) as per the recommendations of Sabari et al. (1998). The patient will be asked to actively move their arm in abduction, forward flexion (abduction + flexion), as well as internal and external rotation with the humerus abducted to 90deg, without assistance from the examiner. A goniometer (baseline goniometer, manufacturer, country) with two adjustable overlapping arms, marked at 1deg increments, will be applied to the humerus (one arm) and to the upper torso at the end of the movement.

Radiographic findings: Radiographs of the humeral head will be assessed with the method described by (Bell and Coghlan 2014) and radiolucency around the glenoid component classified with the Lazarus system (Lazarus et al. 2002). RP has list of references for radiology measures

Complications and reoperations: These will be logged in the research database (history, surgeon exam, follow-up and complications window in Socrates). A list of available complications specific to shoulder arthroplasty can be found below.

Complications and re-operations

Complications specific to arthritis surgery include:

Infection: Surgical site infection will be determined using the diagnostic criteria proposed by Frangiamore et al (2015) and summarised in Table 1 of their paper (Figure 2.2.5).

Nerve injury: The musculocutaneous nerve and dorsal digital nerve of the thumb are particularly at risk during arthroscopy and the brachial plexus is at risk during arthroplasty (Dwyer et al 2015). Neuropathic symptoms may be defined as a complication in patients if they have not resolved after 6 weeks from surgery.

Prosthesis fixation failure/reaction: Component breakage postoperatively and in a symptomatic patient will be logged as a complication. Patient immunoreaction to prosthesis materials (e.g synovitis diagnosed by imaging) will also be logged as a complication.

Deleterious bone remodelling: Diagnosed on radiographs by the treating surgeon as lysis.

2.2.10 Statistical analysis plan

Background

Here the pre-determined statistical analysis plan for the shoulder arthritis cohort is described, which was finalized prior to any data analysis and to which the investigators will adhere in future analyses. The statistical analysis plan was completed and approved by the registry steering committee on [insert date]. Prospective participant recruitment was commenced [insert date] and following data integrity checks (as per Section 3 of this document), the cohort dataset (including retrospective data prior to relaunch) will be locked on an annual basis and the statistical analysis as specified below will be performed.

Analysis Objectives/Questions

The analysis plan is derived directly from Question 1 listed above. Question 2 will be best answered with a prospective randomised trial supported by this registry cohort.

Design

The shoulder arthritis cohort analysis is a non-randomised observational study with census sampling (all patients presenting to the registry sites are included). Consecutive patients from the community seeking care for shoulder arthritis pathologies following primary care referral will be screened by the participating consultant / orthopaedic surgeon for eligibility into the analysis.

Patient inclusion/exclusion

Alterations from cohort criteria: No

Treatment allocation

Patients will be allocated to a treatment pathway (surgery, arthroplasty, non-surgical management, observation-only) as per the consultant surgeon’s clinical judgement and shared decision-making with the patient.

Sample Size

Sample size calculations were based upon a rule of thumb of 8 failure events per variable entered into the model (Vittinghoff and McCulloch 2007), assuming a failure rate of 30%. A sample size of 176 patients will have 80% power to detect a significant effect for 6 variables, allowing for up to 10% cohort protocol non-compliance and a two-sided alpha of 0.05. The failure rate is estimated from Gauci et al (2018), which described clinical outcomes after anatomic TSA in a series of 69 patients.

Preliminaries

The flow of patients through entry into the cohort and analysis will be displayed in a flow diagram as per STROBE guidelines (Strengthening The Reporting of Observational Studies in Epidemiology) (Figure 6) (Vandenbroucke et al. 2007). The number of exclusions and loss to followup will be noted from screening, data collection and analysis phases.

Data collection and follow-up

As per registry methods. Approval is also sought to invite patients who have not previously consented to participate in research to return patient reported outcome measures for the purposes of this study.

Variations from Cohort core dataset: No

Data Integrity

Initial statistical analysis will include examining data consistency and examining outliers in more detail. Where outliers in patient reported outcomes, surgical or clinical information are identified, source material will be examined and transcription errors rectified. Where the source material contains an outlier value, a decision will be made regarding the validity of the value. This will be reported in the published results and the datum excluded from further analysis if required.

As per Section 3 - quality management

Handling missing data

Missing data will be identified during the initial data quality assessment. Missing data identified in the electronic registry database will be investigated in the original source material where available. In patients with missing data who are eligible for follow-up, attempts will be made to contact the patient (mail and phone) to complete patient-reported outcomes. When the number of patients required to achieve 80% follow up at a minimum of 1 year has been achieved, missing data rates will be reported.

Analyse randomness of missing data (check)

Variables

As per cohort CDS

Outcomes

The proportion of patients labelled as “ongoing” in surgical status, having met the criteria for surgical success.

Surgery/management

As per cohort core dataset

Demographics and baseline characteristics

As per cohort core dataset

Analyses

Principles (incl. blinding)

Randomisation not applied to the analyses. Analyses will be performed with de-identified data (single-sided blinding)

Primary Questions

An analysis request will be submitted to the AOA National Joint Registry to extract all patients meeting the above criteria operated on by the principal investigators with a Kaplan-Meier survival analysis to report procedure survival to maximum of 15 years. 95% confidence intervals will be used to compare results between the study cohort and all other procedures logged in the registry with the same implant performed by other surgeons in Australia.

Outlier analysis and normality will be assessed prior to a descriptive analysis of patient-reported outcomes using appropriate measures of central tendency (mean, median) and variability (95% confidence intervals). Patient-reported outcomes will be summarised using descriptive statistics as appropriate to data type (continuous, categorical). Binary logistic regression will be used to determine the association of demographic, anatomical, pathology and management factors to surgical success at a minimum of 1 year follow up.