|Home | About | Journals | Submit | Contact Us | Français|
Since its description by Paul Grammont from Dijon, France, several tens of thousands of reverse total shoulder arthroplasties (RTSA) have been performed for diverse conditions. The purpose of this analysis is to identify the complications of this procedure in the literature and in clinical practice. A total of 240 papers concerning RTSA published between 1996 and 2012 have been identified. Over 80 papers describe complications associated with this type of implant. A list of prostheses satisfying European and US standards, CE and FDA approved, has been produced on the basis of information provided by the manufacturers. Data from the literature do not support a meta-analysis. The inventory of best practices shows excellent results in the short and medium term in specific indications, while the number of complications varies between 10 and 65 % in long-term series. Complications can be classified into (A) non-specific including infections (superficial and deep), phlebitis, haematoma, neurological complications of the suprascapular, radial and axillary nerves and (B) specific complications associated with RTSA including (1) on the glenoid side: intraoperative fracture of the glenoid and acromion, late fracture of the scapula, impingement at the scapular neck (notching), glenoid loosening, dissociation of the glenoid component (snatching of the glenosphere) and fractures of the glenoid baseplate; (2) on the humeral side: metaphyseal deterioration, humeral loosening, instability of the shoulder, stiffness with limitation of external and/or internal rotation; and (3) muscular complications with fatty degeneration of the deltoid. Additionally we have identified specific situations related to the type of implant such as the disassembly of the humeral or the glenoid component, dissociation of the polyethylene humeral plate, dissociation of the metaphysis and osteolysis of the tuberosities. The integration of results from different clinical series is difficult because of the lack of a database and the multitude of implants used.
In 1993 Paul Grammont, Pierre Trouilloud and Emmanuel Baulot from Dijon, France described the original concept of the modern reverse arthroplasty showing that if the rotator cuff cannot be restored, functional recovery of the shoulder can be obtained with a total shoulder prosthesis, medialising the glenohumeral centre of rotation and elongating the remaining deltoid muscle. The early results were published by Baulot et al. in the Belgian Journal of Orthopaedics . Previous concepts published by the Dijon school included a humero-acromial prosthesis (Acromion) (Fig. 1) and a first-generation reverse implant called the “Trumpet” prosthesis (Fig. 2) that were abandoned. The original Delta prosthesis (Fig. 3) was designed by Grammont, Trouilloud, Baulot and Capon with Michel Colombier, a mechanical engineer from Medinov, a French company that developed this gamut of implants. Medinov was later purchased by Landanger and became a part of DePuy in 1999. DePuy introduced the Delta prosthesis in the USA by 2004 and by that time seven manufacturers were producing different reverse shoulder models . Currently there are more than 20 companies marketing reverse shoulder implants and other manufacturers are developing new designs. The differences between the implants are related to the glenoid fixation (oval or circular glenoid baseplate, long or short central peg, locked or unlocked fixation screws); and to the humeral stem and metaphysis (big or small metaphyseal component, cemented or cementless). An important difference relates to the proximal humeral cut that ranges from 125 to 155° in respect of the cervical-diaphyseal angle and between the surface treatments of the different implants. A common fact related to reverse total shoulder arthroplasty (RTSA) is that all the manufacturers are continuously searching for ways to improve and modify the current implants. The object of this review paper is to identify the different complications related directly to this type of surgery (Table 1).
Indications for RTSA include degenerative arthritis associated with irreparable cuff tears (cuff tear arthropathy) and irreparable cuff tears with loss of function in the elderly (pseudoparalytic shoulder) [3–9], aseptic necrosis of the humeral head in the elderly , shoulder reconstruction in rheumatoid arthritis [5, 10], chronic shoulder dislocations in the elderly , reconstruction surgery for tumour [12, 13], revision surgery after failed anatomical or resurfacing arthroplasty [14–19], failed rotator cuff repair with superior escape and reconstruction for comminuted fractures of the proximal humerus in the elderly [20, 21]. Any procedure addressing shoulder reconstruction with RTSA demands the existence of a functional deltoid muscle that makes shoulder mobility possible with this type of implant . The major indication for using RTSA is arthritis associated with massive rotator cuff tears which accounts for approximately 90 % of the cases in different series [2, 5, 6, 8, 9, 22]. The patient’s age is important for the indication, as the implant longevity is not well understood; therefore, most authors do not advocate use of this type of implant for patients under 65 years of age [1, 5, 9, 23–26]. The use of reverse arthroplasty changes the anatomy of the shoulder, translates the centre of rotation, lengthens the arm and provides a better lever for the deltoid muscle (Fig. 4). The proprioception is changed .
Revision shoulder arthroplasty using a reverse implant is associated with higher rates of complications, such as infection, neurological injury and intraoperative fractures. In addition to these problems, the situation may be complicated by significant proximal humeral deficiency, specifically loss of the proximal shaft, metaphysis and tuberosities [9, 16, 18, 28, 29].
The assessment of the results with this type of arthroplasty was made using general and specific tools. The general tools for the clinical assessment of patients treated with RTSA was made with different validated scores and tools such as the simple shoulder test (SST), the Constant and Murley score, the American Shoulder and Elbow Surgeons (ASES) score and the University of California, Los Angeles (UCLA) score. A specific tool for the assessment of glenoid bone loss when a reverse shoulder implant shows notching and lysis of the inferior pillar of the scapula was described by Cécile Nérot and François Sirveaux from France, cited by Lévigne et al. .
The literature relating to RTSA has been abundant lately and includes more than 50 new papers per year. The increase has been exponential during the last 24 months recorded by our research of the literature and there are several papers dealing with complications and eventful outcomes.
Thorough mastery of the procedure of implantation for different indications seems to be crucial as the complications are more frequent during the first implantations, as suggested by Kempton and colleagues. From a series of 200 reverse arthroplasties performed in 191 patients by a single surgeon including 40 revisions the local complication rate was higher in the first 40 shoulders (23.1 %) versus the last 160 shoulders (6.5 %). The authors conclude that the early complication-based learning curve for RTSA is approximately 40 cases. There were more complications and more neuropathies in revision versus primary reverse arthroplasties .
The Delta reverse shoulder was the oldest and the most widely used implant in its three consecutive versions (Delta 1, Delta 2 and Delta Xtend), which is probably why problems with this type of prosthesis are frequently reported in the literature [3, 8, 28, 34–38], but other implants were recorded with similar complications [7, 8, 16, 37, 39–41].
General problems related to the outcome of an arthroplasty include:
A comprehensive list of complications reported in the literature has been published by Wierks et al. . They define early complications such as glenoid intraoperative fracture or unscrewing of the glenosphere.
A study by Molé and Favard from a multicentre European group included 527 patients undergoing RTSA, mainly with a Delta prosthesis. There were 3 % haematomas, 5 % infections, 3.4 % instability, 5 % complications on the glenoid side, 2 % complications on the humeral side, 3 % fractures of the acromion and 1 % neurological complications .
A study by Farshad and Gerber follows two series, a first one of 111 cases of Delta III (DePuy) and a second series of 230 cases of Anatomical Reverse® (Zimmer, Inc.) prostheses. Non-specific complications such as haematoma, infection and neurological impairment are reported, as well as 50 % cases of scapular notching in the case series of 230 patients followed up by Gerber at an average follow-up of 22.3 months. Glenoid component loosening was observed in 3 % of the cases .
Scapular notching is by far the most common situation encountered with RTSA. Its occurrence is dependent on several factors, but it seems to be related to the glenoid morphology because the same implant may result in notching in a short-neck glenoid and no notching in a long-neck glenoid as demonstrated by Pierre Trouilloud and co-workers from Dijon in anatomical experiments. Different solutions have been proposed to prevent notching: Frankle et al. suggested lateralising the centre of rotation outside the scapula, using an eccentric glenosphere that approaches two thirds of a sphere volume instead the one half used in the original Delta concept, creating more constraint and torque on the glenosphere and increasing the risk of glenoid loosening. Increasing the inclination (neck-shaft angle) of the humeral component will avoid inferior scapular notching , but then may create a superior conflict and enhance prosthetic loosening. Baulot and co-workers propose a new glenoid design manufactured by Aston Medical (Duocentric®) with extended inferior coverage that provoked no notching in a case series of 50 patients (unpublished data). A retrospective study compares clinical and radiological results of 47 Delta III reverse prostheses and 49 Arrow reverse prostheses at a minimum of 12 months follow-up. Scapular notching was noted in 32 patients with the Delta III prosthesis and in no instance with the Arrow prosthesis. The design features of this implant were found to be associated with improvement in range of motion and absence of scapular notching .