Reverse total shoulder arthroplasty for irreparable rotator cuff tears without arthritis: A systematic review

Abstract

Introduction

Indications for reverse total shoulder arthroplasty (RTSA) have expanded over recent years. Whilst cuff tear arthropathy is an accepted indication, the results of its use in those without arthritis is not clear. The aim of this article is to review the literature on RTSA for massive rotator cuff tears without associated arthritis.

Methods

A systematic review search was performed in accordance with Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines to find all studies with clinical outcomes on RTSA performed for massive irreparable rotator cuff tears (MIRCT) without arthritis.

Results

Of the 160 studies produced by the search, a detailed analysis found 11 articles to be included in this review. There was variability in the implant style used and the outcome measures utilised, but all studies found improvement in the outcome following RTSA. Many studies advised judicious use following high complication rates, and caution was advised in those patients with pre-operatively preserved active forward elevation.

Conclusion

The available evidence suggests that RTSA is a reliable option in older patients with persistent pain and lack of function following MIRCT even without arthritis. However, as outcomes are not significantly worse following failed rotator cuff repair, joint preserving options in the younger age group should be carefully considered in light of the relatively high complication rate associated with RTSA.

Level of evidence

IV.

1. Introduction

Indications for reverse total shoulder arthroplasty (RTSA) have expanded over the past two decades since its introduction for low demand patients with glenohumeral osteoarthritis and rotator cuff dysfunction.¹ Increasingly, one of the groups of patients that poses a dilemma, is the massive or irreparable rotator cuff tear (MIRCT) without associated glenohumeral arthritis. Defining the MIRCT is also challenging: a massive rotator cuff tear may still be repairable if it is acute, and a moderate sized rotator cuff tear may be irreparable if it is chronic and associated with musculotendinous changes. Commonly accepted definitions of MIRCT include a tear size >5 cm and/or a tear involving ≥2 tendons.^{2, 3, 4, 5} More important than the size of the tear however, is the ability to repair it and in turn, the ability for the repair to heal. This may be based on clinical factors – such as the chronicity of the tear and age of the patient, on imaging appearances – such as degree of tendon retraction,⁶ atrophy or fatty infiltration of muscles,^7,8 or on the inability to repair the tear intraoperatively.

The function of the rotator cuff is to centre the humeral head on the glenoid, and by creating a concavity-compression,⁹ establish a fulcrum around which the deltoid can act to elevate the arm. As such, with an intact cuff, the force-coupling ensures minimal translation of the humeral head on the glenoid throughout abduction elevation. An isolated supraspinatus tear, not extending into the infraspinatus, is also unlikely to cause a large increase in deltoid force, as co-contraction of the large adductors compensate, and superior migration of the humeral head would be unexpected.^10,11 In a postero-superior rotator cuff tear, force coupling is lost; the shear forces of deltoid is no longer opposed by concavity-compression, and the humeral head migrates superiorly.¹² This usually leads to wear of the superior aspect of the glenoid. Gradually osteoarthritis secondary to rotator cuff disease occurs, and cuff tear arthropathy ensues.¹³

If a large rotator cuff tear is irreparable, then even without glenohumeral arthritis, persistent pain and dysfunction may prompt operative intervention. Pain relieving options suggested have been debridement with or without long head of biceps tenotomy,¹⁴ insertion of a balloon spacer¹⁵ and hemiarthroplasty¹⁶ as a limited goal procedure. Other options include partial cuff repair,¹⁷ patch augmentation,^18,19 tendon transfers,²⁰ superior capsular reconstruction (SCR),²¹ and RTSA.²² RTSA for a deficient rotator cuff without overt arthritis is a relatively new indication, and the results of arthroplasty for this indication are not yet clear. The aim of this article is to systematically review the existing literature on the outcomes of RTSA performed for MIRCT without arthritis.

2. Methods

2.1. Search strategy

A systematic review was performed using the Preferred Reporting Items for Systematic reviews and Meta-analysis (PRISMA) guidelines. The search was performed using Web of Science, Cochrane, MEDLINE, Embase, SCIelo and google scholar databases on February 16th^, 2021. A search strategy based on a topic result of “irreparable rotator cuff tear” OR “massive cuff tear” OR “pseudoparalysis” OR “pseudoparesis” OR a title result of “without arthritis” was combined with a title search of “reverse shoulder arthroplasty” OR “reverse shoulder prosthesis”.

2.2. Study selection

All review articles, case reports, animal studies, biomechanical studies, surgical technique reports, and tumoral studies were excluded. Results were screened for the inclusion criteria of 1) RTSA implanted in at least one treatment group, 2) underlying diagnosis of irreparable rotator cuff tear, massive rotator cuff tear, or pseudoparalysis, 3) includes at least one treatment group without arthritis, 4) outlines clear clinical outcomes with analysis of the group without arthritis, 5) details surgical intervention undertaken. Studies not in English, and studies without clinical outcomes were excluded. The study selection is detailed in Fig. 1.

Fig. 1.

Flow chart to show methodology of study selection and reasons for excluded articles.

3. Results

The initial key term search identified 160 articles. 2 duplicate articles were removed, and a title and abstract screening for relevance found 73 articles. Further abstract screening for inclusion criteria eliminated 52 studies, leaving 21 articles. A full-text analysis found 11 studies that met the inclusion criteria.^{23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33} A further 3 studies were found which outlined complications only rather than clinical outcomes.^{34, 35, 36} Table 1 shows the 11 included studies and the study demographics with clinical outcomes measured. Table 2 shows the study descriptions, conclusions and complications, and any declared conflicts of interest. Only 2 studies included declared no conflicts of interest.^30,32

Table 1.

Included studies with study designs and demographic information.

AUTHORS	YEAR	TOTAL NUMBER ANALYSED	ANALYSED RTSA WITHOUT ARTHRITIS	CONSECUTIVE	STUDY DESIGN	LOSS TO F/U	IMPLANT	INCLUSION	AV AGE	EXCLUSION	F/U	OUTCOME MEASURES
ALLERT et al.24	2018	187	92	N	Retrospective – single surgeon	0	DJO	>65y, no OA	74	Patients with arthritis, patients with previous non-arthroscopic surgery	47	Satisfaction, ROM
EK et al.25	2013	46	21	Y	Retrospective review - Single centre	6	Delta III (80%) and Anatomical Shoulder Reverse (20%)	<65y; MIRCT; pseudoparalysis; with or without GHOA	60	<5y f/u, no MIRCT	93	Constant, ROM, strength and pain scores
HARTZLER et al.23	2015	74	74	N	Retrospective case control – single surgeon	0	DJO RSP	MIRCT without OA	72	Any other indication for rTSA	43	SST, ASES
KAAB et al.29	2021	211	53	Y	Prospective observational - multicentre	54	Affinis Inverse	rTSA for CTA or MIRCT who failed non-op treatment or joint preserving surgery	75.8	Not stated	79.9	Constant, ASES, VAS, Satisfaction, ROM
KIM et al.30	2020	98	31	Y	Retrospective – single surgeon	0	Aequalis	All rTSA by senior author	68.9	Less than 24 months f/u	48.4	Constant, UCLA, VAS pain, ROM, satisfaction
LACHETA et al.31	2020	60	33	N	Retrospective cohort – single surgeon	9	Grammont-style	MIRCT with a choice of SCR or rTSA, intact Teres minor and reparable subscapularis, under 70, no OA	63	Tumour, fracture, prev arthroplasty, irreparable subscapualaris, deltoid deficiency, revision SCR or rTSA, ipsilateral upper limb abnormal neurology	2.9	ASES, SANE, QuickDASH, SF-12, satisfaction
LEE et al.32	2020	127	34	N	Retrospective matched cohort - multicentre	0	Equinox (Exactech)	rTSA for either IRCT or CTA	73.5	Prev surgery for fracture, bone erosion needing augmentation, AVN, concomitant NM disease, primary OA/RA	33.2	Constant, ASES, ROM, VAS pain
LINDBLOOM et al.33	2019	699	202	N	Retrospective review – single centre	0	DJO	rTSA for any indication except revision	71	Prev arthroplasty, <24mth f/u	47	ASES, SST, ROM, VAS pain, SF-12
MULIERI et al.26	2010	72	60	N	Retrospective review – single centre	12	DJO RSP	MIRCT without OA	71	Any other indication for rTSA	52	ASES, ROM, SST, VAS pain
WALL et al.28	2007	240	41	Y	Retrospective – single centre	41	Delta III and Aequalis	rTSA for any indication except revision	72.7	rTSA for tumour, acute fracture or rheumatoid arthritis	39.9	Constant score
BOILEAU et al.27	2009	42	25	N	Retrospective - multicentre	0	Delta and Aequalis	rTSA for failed rotator cuff repair with persistent pain and/or loss of elevation	71	Previous arthroplasty, tumour, fracture, rheumatoid	50	Constant, ROM, VAS pain

Table 2.

Studies with design description, conclusions, complications and declared conflicts of interests (OA = osteoarthritis; RCR = rotator cuff repair; MIRCT = massive irreparable rotator cuff tear; IRCT = irreparable rotator cuff tear; GHOA = glenohumeral arthritis; CTA = cuff tear arthropathy; TT = tendon transfer; SCR = superior capsular reconstruction).

AUTHORS	YEAR	TOTAL NUMBER ANALYSED	ANALYSED RTSA WITHOUT ARTHRITIS	STUDY	F/U	OUTCOME MEASURES	CONCLUSION	COMPLICATIONS	LEVEL OF EVIDENCE	COI
ALLERT et al.24	2018	187	92	Retrospectively compared 2 arms – rTSA and RCR	47	Satisfaction, ROM	Both are good options, but those who had rTSA had poorer function pre-operatively	Reoperation 2.2%	IV	Authors receive payment from DJO
EK et al.25	2013	46	21	Retrospectively compared outcomes of rTSA for pseudoparalysis due to MIRCT with or without arthritis, and those with previous surgery to those without	93	Constant, ROM, strength and pain scores	rTSA offers improved function which is maintained up to 10y, but with high complication rate. No difference in functional outcome between those with or without arthritis	37.5% complication, 15% revision; 27.5% re-operation; 56% notching	IV	Author receives royalties from Zimmer
HARTZLER et al.23	2015	74	74	Retrospectively created 2 arms dependent on improvement in SST. Included patients who had lat dorsi TT in combination with rTSA	43	SST, ASES	Age <60, neurologic dysfunction and a pre-op SST >7 were prognostic of poor outcome	12% reoperation, 5% complication	III	Authors receive payment from DJO
KAAB et al.29	2021	211	53	Prospective multicentre study with 2 arms – those with and those without arthritis	79.9	Constant, ASES, VAS, Satisfaction, ROM	Hamada grade does not correlate with clinical outcomes	7.2% complication, 6% revision	III	Funded and stats by Mathys Affinis
KIM et al.30	2020	98	31	Retrospectively compared 2 arms - CTA and MIRCT without arthritis	48.4	Constant, UCLA, VAS pain, ROM, satisfaction	No differences in outcomes regardless of underlying diagnosis	5% complication, 17% notching	IV	None
LACHETA et al.31	2020	60	33	Retrospectively compared 2 arms – MIRCT without arthritis treated with either SCR or rTSA	2.9	ASES, SANE, QuickDASH, SF-12, satisfaction	SCR gives similar outcomes to rTSA in younger patients without GHOA	Reoperation 3%	III	Funded by Arthrex
LEE et al.32	2020	127	34	Retrospectively compared a matched cohort of rTSA done for CTA or done for MIRCT without arthritis	33.2	Constant, ASES, ROM, VAS pain	Significantly improves after rTSA regardless of pre-op diagnosis	1 revision CTA	III	None
LINDBLOOM et al.33	2019	699	202	Retrospective review of all rTSAs and then analysed according to pre-op diagnosis	47	ASES, SST, ROM, VAS pain, SF-12	Males without OA had worse satisfaction; CTA had more predictable results	1% revision	III	Funded by DJO
MULIERI et al.26	2010	72	60	Retrospective review of all rTSA done for MIRCT without arthritis. Compared those who had previous surgery to those who hadn’t	52	ASES, ROM, SST, VAS pain	rTSA a good option in cuff deficiency when other options failed	20% complications	IV	Authors receive payment from DJO
WALL et al.28	2007	240	41	Retrospectively compared rTSA done for CTA and those done for MIRCT without arthritis	39.9	Constant, ROM	rTSA can be used, but worse results in revision and post-traumatic cases, suggest caution judicious use. Best outcomes in primary OA, MIRCT without arthritis and CTA	3.5% removal or revision, 19.1% complication	II	Authors receive payment from Tornier
BOILEAU et al.27	2009	42	25	Retrospectively compared those with pseudoparalysis to those without, also analysed results of those with and without GHOA	50	Constant, ROM, VAS pain	Active anterior elevation decreases in those who had >90° pre-op	12% complication, 5% reoperation	IV	Authors receive payment from Tornier

In total, 666 patients with MIRCT without arthritis had RTSA as their treatment. A lateralised design implant was used in 462 subjects and 204 had a medialised Grammont-style prosthesis. 10 of the 11 studies were retrospective reviews or retrospectively created cohorts. Only one study provided level II evidence, the rest were level III or IV. 9 of the 11 studies declared conflicts of interest. All except one study gave details on complications of the procedures.³²

8 studies detailed the range of movement (ROM) post-operatively as an outcome measure. On average, the post-operative active elevation achieved after RTSA was 135°. Fig. 2 shows the average change in active elevation achieved with the implant design style for each study. The difference between the average change in active elevation between the lateralised designs (64.7°) and the medialised designs (52.07°) did not reach statistical significance.

Fig. 2.

Average change in active elevation achieved for each study (implant design style under first author)

6 studies used the Constant-Murley score³⁷ as a post-operative outcome measure, with the average post-op score being 63.4. The average change in Constant score for patients with MIRCT without arthritis was 33 points. All six studies had full pre- and post-op Constant scores for those patients who were included in the post-op analysis.

Complications following the procedure were documented variably, with some papers including grade 1 notching as a complication and others only mentioning revision procedures as complications. There were 3 studies that analysed complications as their only outcome measure – these are shown in Table 3. One study looking at RTSA in the under 60y age group, found complications seen in the RTSA group without arthritis was 56% compared with 29% in the group with arthritis.³⁴ Another study showed that notching was a problem in RTSAs performed for pseudoparalysis following MIRCT with concurrent arthritis compared with those without arthritis, with rates of 74% and 58% respectively.³⁶

Table 3.

Studies only comparing complications between the arthritis and non-arthritis cohorts (MIRCT = massive irreparable rotator cuff tear).

AUTHOR	YEAR	TOTAL NUMBER OF SUBJECTS	TOTAL NUMBER OF RTSA FOR MIRCT WITHOUT ARTHRITIS	CONSECUTIVE	STUDY DESIGN	IMPLANT USED	INDICATION	AV AGE	FOLLOW UP	CONCLUSION	COMPLICATIONS	COMMENTS
ERNSTBRUNNER et al.34	2017	26	9	Y	Retrospective Single centre	Delta III/Anatomical shoulder reverse	rTSA for MIRCT	57	11.7y	rTSA in under 60y gives improvement in function without clinical deterioration beyond 10y	39% overall complications	Complications in group without arthritis was 56%, compared with 29% in the group with arthritis
GERBER et al.35	2018	22	11	Y	Longitudinal observational Single centre	Delta III	rTSA for irreparable cuff and pseudoparalysis	68	16.1y	High complication rate, but in those not needing explantation, still have good functional outcome	59% complications, 55% reoperation, 27% failed	No difference in complications between the group with and the group without arthritis
KATZ et al.36	2016	140	39	N	Retrospective Single surgeon	Arrow lateralised	Pseudoparalysis due to MIRCT c/s OA	72	3.75y	Metallic lateralisation resulted in lower scapular notching rates	29% notching - higher if BMI lower; more Hamada 4–5 in notched group; 8.9% revision	74% scapular notching seen in group with arthritis compared with 58% in group without

4. Discussion

The designs of RTSAs have changed over the years to provide better glenoid baseplate fixation and to provide options of medialised or lateralised centre of rotation. In addition, surgical experience has increased and consequently indications for RTSA have expanded. Primary osteoarthritis, cuff tear arthropathy (CTA) and inflammatory arthritis remain recognised mainstay indications, but RTSA being performed for trauma (both acute and sequelae) and for MIRCT without arthritis have increased over the past few years.³⁸

One of the problems with MIRCT without arthritis is lack of function. Pseudoparalysis has had inconsistency in its definition,³⁹ but most studies included in this review use the definition of active forward elevation <90° with a maintained passive ROM. This has been highlighted as a key pre-operative consideration in four studies, with Boileau et al.²⁷ finding a decrease in active elevation of 24° in those with preserved movement. Hartzler et al.²³ also allude to poor outcomes in this group, although they looked at all independent risk factors for a non-clinically important difference in increase in simple shoulder test (SST) scores and found pre-op SST ≥ 7, as well as age <60y to be relevant.

MIRCT can also occur in a younger age group, and one study looking at RTSA outcomes in the under 65s found no significant deterioration of relative Constant score or shoulder satisfaction over 10 y²⁵. In this study of 41 patients – 21 without arthritis – 6 implant failures were excluded from their final analysis, and in total 11 cases required reoperation (27.5%). Lacheta et al.³¹ compared SCR to RTSA in the under 70y age group and found that of the 28 who played sport pre-operatively, 96% were able to return to sport – there were no significant differences between the study groups. In each group of 22 and 29 respectively, 1 patient required revision. Hartzler et al.²³ studied a group of 74 patients who underwent RTSA for MIRCT without arthritis and identified age <60y at time of operation to be an independent risk factor for poor outcome. Interestingly, they did not find a major post-operative complication to independently affect outcome. Ernstbrunner et al. looked at complications in those under age 60 and compared rates between those with and without arthritis.³⁴ Notching was found in 95% which increased over time, and in their cohort of 23 patients with an average age of 57, they found a clinical complication rate of 39% (56% in those without arthritis).

The complication rates in RTSA are certainly a reason why joint preserving options are appealing to use in the younger age group. In their systematic review, Kooistra et al. identify all treatment modalities for MIRCT and found that SCR had the best post-operative improvement in Constant score.⁴³ However, it is a new procedure and surgical experience is not yet at the level of that for RTSA. The indications are very specific²¹ and obtaining predictable results remains a challenge.

Variability in patient functional demands, pre-operative active elevation and chronicity of the rotator cuff tear are all important aspects when evaluating a patient for surgical treatment for MIRCT without arthritis. The largest study included in this review found worse satisfaction rates in men undergoing RTSA for MIRCT without arthritis, and more predictable results in RTSA for CTA.³³ However, most studies found a high complication rate with RTSA, and did not find underlying diagnosis to affect this.

5. Conclusion

In conclusion, the available evidence suggests that RTSA is a reliable option in older patients with persistent pain and lack of function following MIRCT even without arthritis. However more than one study has shown age under 60 to be an independent adverse risk factor, and RTSA still has a relatively high complication rate. Given that results of RTSA done for failed rotator cuff repair and those done as a primary RTSA are not significantly different,^24,26 it may be suggested that joint-preserving options are considered first-line for those younger patients with MIRCT without concurrent arthritis.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.