Re-tears after rotator cuff repair: Current concepts review

Abstract

Re-tear following rotator cuff repair is common and has been reported to range from between 13 and 94% despite satisfactory clinical outcomes following rotator cuff surgery. Various risk factors have been associated with an increased tear rate, including patient factors, tear and shoulder morphology, repair technique, and rehabilitation regimes. Different modes of rotator cuff failure have been described. The management of re-tear in patients following rotator cuff repair is challenging and depends on the age, functional status and requirements of the patient, and re-tear size and residual tendon length. This article aims to review the factors associated with rotator cuff re-tear. It describes which of these are associated with poor clinical outcomes, and discusses the long-term outcomes of re-tear and treatment options.

1. Background

Re-tear following rotator cuff repair is common and has been reported to range from between 13 and 94%¹ despite satisfactory clinical outcomes following rotator cuff surgery.^2,3 Various risk factors have been associated with an increased tear rate, including patient factors, tear and shoulder morphology, repair technique, and rehabilitation regimes.4, 5, 6, 7, 8, 9, 10, 11, 12 Rotator cuff re-tear is commonly assessed using the MRI-based Sugaya classification,¹³ but ultrasound and CT arthrogram are common techniques to evaluate cuff integrity following surgical repair. Different modes of rotator cuff failure have been described. The most common of which is a failure of the tendon medial to the repair site and less commonly through biomechanical failure. The management of re-tear in patients following rotator cuff repair is challenging and is dependent on the age, functional status and requirements of the patient, as well as re-tear size and residual tendon length. This article aims to review the factors associated with rotator cuff re-tear and which of these are associated with poor clinical outcomes, and to discuss the long-term outcomes of re-tear and the treatment options.

2. Risk factors for rotator cuff re-tear

2.1. Patient factors

Increasing age has been associated with increased rates of re-tear following rotator cuff repair.^4,7 Diebold et al. described a large single surgeon series of 1600 patients and identified a re-tear rate of 5% in patients under the age of 50 years, 10% re-tear rate in patients aged 50–59 years, 15% in patients aged 60–69 years and 25% in patients aged 70–79 years and increasing to 34% in patients over the age of 80 years.⁴

Systemic factors such as low bone mineral density (BMD), hyperlipidaemia (HL) and poor glycaemic control in diabetic patients have also been shown to be significant risk factors associated with failure of rotator cuff repair.14, 15, 16, 17, 18 Chung et al.¹⁴ demonstrated that patients with osteoporosis (BMD < -2.5) had 7.25 times the probability of failure of rotator cuff repair, and those patients with osteopaenia (BMD between −2.5 and −1) had a 4.38 times the probability than patients with normal BMD. BMD remained a significant independent risk factor with multivariate logistic regression analysis, taking into account age, gender, diabetes, tear size and muscle fatty infiltration. HL, has been shown both in animal models and clinical studies to increase the risk of rotator cuff disease and re-tear after rotator cuff surgery.^15,16,19 A major confounder is the use of statin medications. A study by Garcia et al.¹⁵ demonstrated a 6.5 times higher probability of re-tear following rotator cuff repair than patients with normal lipid levels with no difference found with statin dosage or type of medication. Despite BMD, HL and glycaemic control being modifiable risk factors, it is still unclear whether correction of these variables will reduce the risk of failure following rotator cuff repair.

The effect of a lateral downsloping acromion measured by the critical shoulder angle (CSA) as shown in Fig. 1, has been shown to increase the load to the supraspinatus tendon during abduction.²⁰ Clinical studies have also demonstrated an increased risk of re-tear after rotator cuff surgery. A study by Garcia et al. demonstrated those patients with a CSA>38° have an odds ratio of 14.8 of having a full thickness re-tear.²¹ Reducing the CSA by performing a lateral acromioplasty has been shown to be safe and may reduce the risk of re-tear after rotator cuff repair.²²

Fig. 1.

Shoulder X-Ray measuring critical shoulder angle (CSA) of 37°.

2.2. Tear morphology

Multiple tendon variables have been investigated in the role of tendon failure after rotator cuff repair. Size of the rotator cuff tear both in mediolateral dimension, specifically medial to the musculotendinous junction²³ and anteroposterior dimensions and tear size area have been correlated with the risk of re-tear at 6 months. Tashjian et al.²⁴ reported on a series of 51 cases and identified healing rates of 92% in patients who had an initial tear lateral to the musculotendinous junction compared with 56% when the pre-operative tear was medial to the musculotendinous junction.25, 26, 27 Park et al.²⁸ reported outcomes of 339 patients undergoing arthroscopic repair of small to medium-sized tears and found healing rates in tears <2 cm of 89% and tears ≥2 cm had a healing rate of 65%. Studies reporting the intra-operative quality of the tendon, tendon mobility and fatty infiltration of the muscle on pre-operative evaluation by MRI,²⁹ as shown in Fig. 2, or CT⁶ have also been shown to affect the re-tear rate.^6,9,25

Fig. 2.

MRI Sagittal PD image showing severe atrophy and fatty infiltration of the supraspinatus muscle.

Fatty infiltration of the muscle belly has been shown to be an irreversible change correlated with a higher incidence of failure of rotator cuff repair.⁶ Infraspinatus changes of grade II or higher are correlated with an increased incidence of rotator cuff re-tear.²⁸

2.3. Repair technique

Open, arthroscopic single row (SR), double row (DR), and trans-osseous equivalent (TOE), also known as suture bridge (SB) techniques of rotator cuff repair, have all been described. Rotator cuff healing is dependent on providing adequate mechanical stability of the repair until tendon to bone healing is achieved. Double row and TOE repairs have improved biomechanical performance over single row techniques, with many biomechanical studies demonstrating improved tensile strength, load to failure, footprint contact area, footprint pressure and superior resistance to gap formation. The development of TOE repair techniques was developed to address the issues of tendon strangulation, impaired tendon vascularity and load distribution between anchors. A systematic review assessing 2048 repairs by Hein et al.³⁰ found significantly lower re-tear rates in double row and suture bridge techniques compared with single row techniques for most tear sizes, including tears between 1 and 3 cm, less than 3 cm, greater than 3 cm and greater than 5 cm.³⁰ There is some limited evidence that TOE techniques may reduce the risk of full thickness re-tears in patients with large and massive tears especially for tears >3 cm. However, clinical outcomes are not consistently superior when comparing repair techniques.²³ In small and medium-sized tears, most studies showed similar clinical results between double row and TOE techniques.^13,23,31,32 Predictive algorithms have been developed based on these risk factors and are based on anteroposterior tear size and age being more heavily weighted in providing a percentage risk of re-tear.^18,25 A summary of risk factors and their relative effects on re-tear rates are summarised in Table 1.

Table 1.

Risk factors for rotator cuff re-tear.

Risk factors	Effect
Patient factors
Age4	Tear rates
<50 years	5%
50–59 years	10%
60–69 years	15%
70–79 years	25%
>80 years	34%
BMD14	Relative risk
Normal (T-score > −1)	1
Osteopaenia (−1 > T-score > −2.5)	4.38
Osteoporosis (−2.5 > T-score)	7.25
Hyperlipidaemia (HL)15	Relative risk
Normal lipid levels	1
HL	6.5
Critical shoulder angle (CSA)21>38°	Odds ratio 14.8
Tear morphology
Tear size28	Healing rates
<2 cm	89%
≥2 cm	65%
Repair technique30
Double row (DR)	Improved biomechanical performance and lower retear
Trans-osseous equivalent (TOE)/Suture bridge (SB)	rates in tears >1 cm in DR and TOE compared with SR
Single row (SR)

2.4. Rehabilitation

A systematic review of meta-analyses by Saltzman et al.¹¹ identified no difference in re-tear rates between early and delayed motion after cuff repair. Early motion was considered as an active range of motion within six weeks of surgery or passive range of motion within three weeks, and a delayed motion was considered as active or passive range of motion starting at least three weeks from the date of surgery. A sub-group analysis suggested that larger tears >3 cm may have significantly high re-tear rates with early-motion rehabilitation.¹¹ In general, a period of immobilisation of 2–4 weeks, depending on tear size, is recommended. A graded supervised rehabilitation program is beneficial in tendon healing and minimises stiffness compared with an early unprotected range of motion.^11,23,33

2.5. Failure mechanism and time to failure

Understanding failure mechanism and timing after rotator cuff repair can help identify which risk factors, repair techniques and rehabilitation methods are relevant for each patient. Failed rotator cuff repair can relate to mechanical failure of the repair construct such as anchor pull out or knot failure, failure of biological healing of the tendon footprint or re-tear of a healed repair such as with medial cuff failure. MRI findings of re-tear have been classified by Cho et al.³⁴ as Type 1, with no remaining tendon at the greater tuberosity footprint or Type 2 where remnant cuff tissue remained at the insertion site as shown in Fig. 3. Type 1 failure can be related to both mechanical failure or failure of biological healing, whereas Type 2 failure implies re-tear of a healed repair. Sugaya et al.¹³ further classified the structural integrity of rotator cuff repairs using T2 -weighted MRI. In Type I findings, the repaired cuff was normal thickness and demonstrated a homogenous low intensity signal. Type II demonstrated normal thickness with a partial high intensity area. Type III demonstrated less than half the thickness compared to normal without discontinuity, suggesting a partial thickness delaminated tear. Type IV demonstrated minor discontinuity suggesting a small full-thickness tear, and Type V represented medium or large full-thickness tears with major discontinuity observed on both coronal and sagittal slices.

Fig. 3.

MRI findings of failed rotator cuff repairs (A) Type 1 failure with no residual tendon attached to tuberosity footprint with major discontinuity, (B) Type 2 failure with tendon stump attached to footprint and medial tendon failure and major discontinuity.

Studies investigating failure mechanisms in single row repairs tend to show Type 1 failure, but in both double row and suture bridge repair techniques, the identified weak point is the medial tendon showing a higher rate of Type 2 failure.^35,36 A study of rotator cuff healing in suture bridge repairs identified that re-tear rates were not related to initial tear size. The percentage of type 1 re-tears after the suture bridge technique increased with the severity of pre-operative fatty infiltration and muscle atrophy but did not affect the rate after single-row fixation.^34,37 These findings suggest that the presence of fatty infiltration may affect the biological healing capacity of repaired tendons.

Prospective imaging studies with longitudinal follow up have identified that the majority (94.7%) of re-tears occur from 6 to 26 weeks following surgery, with a mean time to re-tear being 19.2 weeks.³⁸ A meta-analysis of studies investigating re-tear rates and timing identified that re-tear rates for medium to large tears increase to at least 10–15 months following surgery. Retear rates for massive tears are more variable and may follow a different time course.³⁹

3. Investigation

Imaging of the rotator cuff after surgical repair can be challenging due to imaging artefact from retained anchor material and post-operative changes that make it challenging to assess repair integrity. MRI, ultrasound and CT arthrogram have all been described with varying degrees of sensitivity and specificity. Ultrasound has a 91% sensitivity, 86% specificity and 89% accuracy compared with the intra-operative arthroscopic assessment but is operator dependent.⁴⁰ Detection of re-tear or failed rotator cuff repair using MRI ranges from 70 to 90% and may be improved with arthrography techniques.⁴¹ MRI allows evaluation of the muscle belly, tendon length and quality and other intra-articular structures that can cause symptoms. For these reasons, MRI is the investigation of choice to evaluate a suspected re-tear following rotator cuff repair.⁴²

4. Patient outcomes and natural history after re-tear

Many studies have shown that functional outcomes correlate to post-operative repair integrity. However, many patients have a satisfactory outcome despite the structural failure of the repair.1, 2, 3^{,14,17,26,43,44,44} Reduction in forward elevation strength by 45% and external rotation strength by 42% can be seen after structural failure despite a well preserved overall range of motion.⁴⁴

If clinical assessment and imaging studies suggest a re-tear after rotator cuff repair, it remains important for the surgeon to consider other shoulder or distant pathology as a cause of symptoms. Common concurrent pathologies include adhesive capsulitis, acromioclavicular joint arthritis, long head of biceps pathology, glenohumeral arthritis, suprascapular nerve lesions and cervical spine pathology. Rates of stiffness following rotator cuff repair have been reported between 4.9% and 32.7%.^45,46 Treatment options include physical therapy, hydrodilatation and arthroscopic capsular release. The results of the latter can be effective in improving function and range of motion. Infection is a relatively rare complication following arthroscopic rotator cuff repair with an incidence of 0.27%–1.94%. The most common infecting organism is Cutibacterium acnes or Staphylococcus. Management includes removing all foreign suture and anchor material and multiple surgical debridements together with an appropriate targeted antibiotic course of treatment.

After excluding other causes of poor post-operative outcome, identifying factors that may predict a poor functional outcome after re-tear can help surgeons identify which patients may require further surgery. Lee et al.⁸ identified that a relative change in AP re-tear size was the most influential independent predictor of symptomatic failed rotator cuff repair leading to revision surgery. Kim et al.⁴⁷ identified that young age <65, lower education levels and worker's compensation claims were independent predictors of poorer functional outcomes and lower education and workers compensation were independent predictors of lower satisfaction after recurrent rotator cuff tear.

Despite the structural failure of the repair, long-term outcomes have shown that most patients can return to their original occupation and were able to maintain improvements in daily and recreational activity out to 7 years, compared to their pre-operative status.¹ Another long term outcome study of patients with structural failure after arthroscopic rotator cuff repair by Paxton et al.⁴³ did not identify deterioration in patient-reported shoulder function and pain between 2 and 10 year follow-up despite a high rate of progression to degenerative joint changes.

5. Management of re-tear after rotator cuff repair

5.1. Conservative management

Patients with poor prognostic factors for healing of a revision rotator cuff repair or low demand individuals should trial a course of non-operative management. This should focus on restoring range of motion and strengthening the residual intact rotator cuff, deltoid and periscapular muscles (See Table 2).

Table 2.

Management options after failed rotator cuff repair.

Management	Indications	Aims and principles
Conservative	Irreparable tear Low demand patients	Physiotherapy to optimise strength of residual rotator cuff, deltoid and periscapular muscles
Revision rotator cuff repair	Adequate tendon length High demand patients Male	Correct reversible risk factors Adequate visualisation Tendon mobilisation Tension free repair Strong biomechanical fixation
Superior capsular reconstruction	Male High demand patient Functional subscapularis Preserved acromio-humeral distance Lower BMI Preserved forward flexion	Use of stiff thick graft Restore posterior continuity to infraspinatus tendon to restore force couple
Tendon transfer	Active patient Pain Functional loss of strength	Restore shoulder kinematics and improve function, strength and pain Use of expendable donor muscle-tendon unit Donor and recipient to have similar excursion and tension Donor and recipient to have similar line of pull Donor muscle to replace single function of recipient muscle
Reverse shoulder arthroplasty	Cuff tear arthropathy Pseudoparalysis of the shoulder	Restore pain free functional motion

5.2. Revision rotator cuff repair

Identifying the cause of failure and correcting any reversible factors are important in achieving a successful outcome of revision rotator cuff repair is to be attempted. It is important for the patient and surgeon to understand expectations of surgery and match this to the likely outcomes. The functional and clinical outcomes are satisfactory but still inferior to primary rotator cuff repair. Both open and arthroscopic revision rotator cuff repairs have been shown to improve range of motion, pain and ASES scores. A systematic review by Brochin et al.⁴⁸ reported an improvement in forward elevation from 110° ± 22°–136° ± 18° (P = 0.004). Although external rotation and internal rotation range improved, these findings were not significant. Functional scores also improved following revision surgery but did not reach significance.⁴⁸ The risk of failure is twice that of primary repair showing re-tear rates of 40% (range 0–62%).^49,50 Challenges include poor tendon quality including short residual tendon length, tissue adhesions, progressive fatty infiltration or muscle atrophy, limited footprint area for revision fixation and local tuberosity osteoporosis. The most consistent factor predicting better outcomes identified pre-operative forward flexion with variable cut offs ranging from greater than 90° to greater than 140°. A strong correlation between pre-operative forward flexion and post-operative forward flexion and ASES scores exist. Other factors included pre-operative external rotation range, pre-operative pain scores and pre-operative ASES scores, which all have a moderate correlation with their respective post-operative scores. Male sex is also associated with better outcomes. In the same study of 807 revisions, the overall complication rate was 12%, 8% for open and 16% for arthroscopic surgery. The most frequent complication was a failure of the repair (88%) followed by persistent shoulder stiffness (5%) then infection (4%). Complications necessitating revision surgery were identified in 7/357 (2%) after open surgery and 30/452 (6.6%) after arthroscopic surgery. Reoperations included revision rotator cuff repair, reverse total shoulder arthroplasty, debridement, arthrodesis, subacromial decompression, capsular release, biceps tenodesis, acromioclavicular joint excision and pectoralis major transfer. Contraindications for revision is an infection, and contraindication to arthroscopic techniques include patients with disruption of the deltoid origin.^48,50,51

Revision rotator cuff repair can be challenging due to operative and patient factors. The surgical aims of any rotator cuff repair include adequate visualisation and tendon mobilisation to achieve a tension-free repair with strong biomechanical fixation to achieve biologic healing to the tuberosity. In patients with an irreparable tear, a partial repair focussing on the subscapularis and infraspinatus aims to restore the shoulder force couple to prevent proximal humeral migration. The challenges in revision surgery are at each of these steps. Adequate visualisation and tendon mobilisation may require multiple portals to allow the division of adhesions. For posterosuperior tears adhesions between the tendon and the acromion, the deep part of the scapular spine and distal clavicle requires a release. Subscapularis tears are often retracted and adhered to the undersurface of the coracoid and its base and the glenoid neck and require mobilisation from these surfaces. Supraspinatus tendon mobilisation may also require an interval slide for adequate tendon excursion identify the tear pattern as a crescent, U-, V-, L- or reverse L- shaped tear to allow planning of a tension-free repair. Medialisation of the footprint with single-row fixation may be required to minimise tension, and one can safely perform <10 mm without significant reduction in glenohumeral joint motion.⁵² Tuberosity management can be difficult, and the decision to retain or remove suture anchors needs to be considered. The techniques of suture stacking using a tap to avoid anchor damage, replacement using larger anchors or using trans-osseous suture fixation without anchors can help achieve satisfactory fixation to the footprint. Patients with bone loss from cyst formation in the tuberosity can benefit from bone grafting, the use of suture plates or distal lateral suture anchor fixation.

Optimising biologic healing of the tendon to bone is also challenging in revision surgery. The use of cellular biologic augmentation using platelet-rich plasma, platelet-rich fibrin matrix (PRFM), autologous mesenchymal stem cells (MSCs) and biologic patches are all options but limited evidence exists to support their use.⁵³ Mechanical biologic augmentation involves using xenograft, synthetic or allograft patches to augment the repair in patients with poor tissue quality or to bridge gaps when the tendon cannot be reduced to the footprint. Biomechanical and clinical studies support the use of acellular dermal allograft.^54,55 Larger studies with longer term follow-up are necessary to prove their efficacy and define their place in the management of revision rotator cuff repairs.

5.3. Superior capsular reconstruction

The use of superior capsular reconstruction was popularised by Mihata et al.⁵⁶ for the treatment of massive irreparable rotator cuff tears with good functional outcomes, including improved active elevation from 84° to 148° (P < 0.001), external rotation from 26° to 40° (P < 0.01), acromio-humeral distance increase and improvements in ASES scores from 3.5 to 92.9 (P < 0.0001) and no progression of osteoarthritis or rotator cuff muscle atrophy.⁵⁶ Autologous fascia lata and acellular dermal allograft have been used as graft choice. A more recent follow up of 54 patients with a minimum one year follow up identified a 20.4% failure rate. Factors that were significantly associated with failure included female sex and the presence of a subscapularis tear. Other factors associated with failure included a trend towards greater body mass index (BMI), lower pre-operative forward flexion and lower acromio-humeral distance (AHD).⁵⁷ Patient selection and technical factors using a stiff and thick graft with appropriate care given to restoring posterior continuity to the infraspinatus and posterior capsule may be necessary for achieving good outcomes.

5.4. Tendon transfer

Retear after rotator cuff repair can lead to a massive irreparable tear of the rotator cuff. Disruption of the normal force couples can lead to a decentering of the humeral head on the glenoid resulting in abnormal wear and an increased risk of arthritis. Tendon transfers around the shoulder aim to restore shoulder kinematics and improve function, strength and pain.⁵⁴ Indications for tendon transfer are an active patient with pain and functional loss of strength. Decision making around the choice of tendon transfer should respect these underlying principles: 1) Expendable donor muscle-tendon unit, 2) Donor and recipient muscle should have similar excursion and tension, 3) Donor and recipient muscle should have a similar line of pull, 4) Donor muscle to replace the single function of recipient muscle.⁵⁸

Tendon transfer options for massive irreparable rotator cuff tears are summarised in Table 3.

Table 3.

Tendon transfer options for irreparable rotator cuff tears.

Rotator cuff deficiency	Tendon transfer	Outcomes	Considerations
Posterosuperior	Latissimus Dorsi	Gerber et al.(59) 10-year results show improvements in SSV, Constant score, ROM and strength but slight increase in osteoarthritic change	Intact subscapularis(59) Passive forward flexion and abduction >80°(60) No pseudoparalysis Increased risk of failure in Hamada III or IVa(61) Poor outcomes with Teres minor fatty infiltration Goutallier 3 or 4(62) Change in function of donor tendon from internal to external shoulder rotator
	Lower trapezius with achilles allograft	Elhassan et al.(63) Almost 4 year results show 97% patients with improvements in pain, SSV and ROM.	Anatomic line of pull(64)
	Teres major	Kolk et al.(65) Mean follow up at 10 years showed improvement in SSV, Constant score and pain
Anterosuperior	Pectoralis major	Ernstbrunner et al.(66) 20 year follow up showed significant improvements in Constant score and SSV. Active forward elevation improved but active internal and external rotation reduced from short term follow up. 19% incidence of osteoarthritis	Different line of pull of pectoralis major to subscapularis(67) Better outcomes with centred humeral head and intact supraspinatus(67)
	Latissimus Dorsi +/− Teres major	Elhassan et al.(68) Early results at mean 13 month follow up show significant improvement in pain, SSV, Constant score and ROM	Anatomic line of pull(69)

5.5. Reverse total shoulder arthroplasty (RTSA)

RTSA is a salvage procedure for failed symptomatic rotator cuff repairs. Indications for reverse shoulder arthroplasty (RTSA) are cuff tear arthropathy and pseudoparalysis with a massive cuff tear with or without arthritis. Good deltoid function is required for success after RTSA and is contraindicated in the presence of active infection. Excellent clinical outcomes can be achieved⁵⁹ with 93% prosthesis survivorship at ten years.⁶⁰

6. Summary

Re-tear after rotator cuff repair presents a challenging problem. Understanding the reasons for failure, appropriate patient selection, correcting modifiable risk factors, and a systematic approach to revision rotator cuff repair can lead to satisfactory outcomes. Understanding the other treatment options, indications for use, and their results, including tendon transfers and mechanical augmentation, is useful in patients where revision rotator cuff repair is not indicated.

Declaration of competing interest

None.