Relationship between clinical and surgical findings and reparability of large and massive rotator cuff tears: a longitudinal study

There are many factors that affect recovery following rotator cuff repair. Among those are the patient’s demographics, pre-operative level of disability, clinical presentation, imaging such as humeral head position, fatty infiltration and tear size and surgical findings such as tear size and shape and tendon quality and retraction. The relationship between size of the rotator cuff tear and recovery based on patient oriented outcomes remains controversial. While some authors point out that tear size is not an important contributor to outcomes or overall satisfaction with surgery [1‐6], others [7‐10] report that larger tears have a less predictable recovery of strength and function and are associated with more residual disability.

The relationship between tear reparability and recovery is also a subject of debate. The limited literature has shown that partial repairs, sometimes referred to as functional repairs, provide surprisingly good results despite incomplete closure of the defect of the cuff [11‐15]. The concept of functional repair was first introduced by Burkhart and colleagues in the early 1990’s [2, 16‐18]. The authors proposed that partial repairs would restore the force couple of the humeral head and increase acromion-humeral distance, resulting in dramatic changes in pain and function [16‐19]. Consistent reports of successful results of partial repairs is promising but the failure to demonstrate superiority of complete repair over partial repair in the limited literature might be due to lack of statistical power in studies that have used small sample sizes of patients [13]. Improved life style of the aging population over the past decades demands better function in these individuals, which increases the need for managing larger and more disabling rotator cuff tears. Therefore, investigating factors that affect reparability of a cuff tear and their interaction with disability can help clinicians to choose surgical candidates more accurately.

The purpose of this study was therefore twofold, 1) to examine clinical outcomes following complete or partial repair of large or massive full-thickness rotator cuff tear, and 2) to explore the value of clinical and surgical factors in predicting reparability. We hypothesized significant improvement in pain and function in both partial and complete repair groups. The second hypothesis was that pre-operative active external rotation, strength, and intra operative findings such as size, shape and tendon quality would affect reparability.

This was a secondary data analysis of consecutive patients with large or massive rotator cuff tears who required surgical treatment (full or partial repair) and were followed up for two years. Patients with isolated subscapularis tears were excluded. Standardized clinical and surgical data and disability scores had been collected prospectively and entered into a shoulder research database. All patients were immobilized in an Ultrasling for 6 weeks and followed a standardized rehabilitation protocol. Patients had provided written informed consent for participation in previous studies. Approval for using the existent database was obtained from the Research Ethics Board of the Sunnybrook Health Sciences Centre.

One hundred and twenty two patients (41 women and 81 men, mean age 64, SD = 9, range 40-90) were included in the analysis. There were 86 large (39 fully reparable, 47 partially reparable) and 36 massive (10 fully reparable, 26 partially reparable) tears. Overall, 49 patients had a full repair and 73 had a partial repair. There were 26 crescent-shaped, 26 L-shaped (18 L-shaped, 8 reversed), and 70 U-shaped tears. In the complete repair group, 23 (45%) had an anatomical repair and 26 (53%) had less than 1 cm lateral defect.

Our study showed that higher subjective disability, limited active external rotation, poor tendon quality, larger tear size, and U-shaped tears were more associated with a partial repair. However, both partial and complete repairs had a statistically significant improvement in their disability level and range of motion. This is consistent with the early theories about rotator cuff pathology and its management proposed by Codman some 80 years ago [30]. In his work, he mentioned the potential benefit of partial coverage of the rotator cuff defect. In the early 1990s, Burkhart [16‐18] coined the term functional tear of rotator cuff to describe a tear that was anatomically deficient but biomechanically intact. He hypothesized that by improving the transverse force couple between the anterior cuff (subscapularis) and the posterior cuff (infraspinatus and teres minor), the overall function of the shoulder joint can improve.

The limited literature [11‐15, 31, 32] in this area has reported promising results for partial repairs of the rotator cuff which is consistent with our findings. In our study, the partial repair group showed a statistically significant improvement in range of motion, strength and disability level over a period of two years. Having a more obvious improvement in external rotation in the partial repair group is a new finding that supports Burkhart’s theory of improving the transverse force by providing more coverage of the cuff defect. Despite the dramatic improvement in clinical and disability scores, the partial repair group appeared slightly more disabled at 2 years when scores were compared cross-sectionally (post-op scores compared at one point). Similarly, comparison of the amount of change (post-op minus pre-op scores) showed less favorable results in the partial repair group in strength, CMS and ShortWORC scores.

A statistically significant improvement in external rotation of the partial repair group is a positive finding which is potentially related to a more significant weakness in external rotators. Wellmann et al, who examined a small sample of 24 patients with a large retracted rotator cuff tear found improvement in active flexion and abduction with a slight deterioration in external rotation [5]. This inconsistency may be related to differences in patient demographics, tear size or surgical techniques. However, since active external rotation appears to be a predictor of reparability, further examination of its significance is warranted.

Two studies have compared the outcomes of complete repair with partial repair [12, 13, 33]. Iagulli et al compared 45 patients with a complete repair to 41 patients with a partial repair and found no statistically significant difference in their UCLA at 2 years [13]. Kim et al reported similar outcomes between healed and unhealed groups of patients with a partial repair [33]. Another study [12] of 41 non-randomized patients with a partial repair (margin convergence) to a group of patients with a repair of the posterior interval slide edge showed no significant difference in range of motion and patient-related outcomes between groups. Superior results with a complete repair is expected as a complete repair is associated with better coverage of the humeral head and therefore better pain relief and improved function and the lack of difference found in the previous studies is potentially due to a lack of statistical power or type of outcome measure used.

In terms of clinical findings, the massive tear group showed a higher incidence of pseudoparalysis. However, active flexion of less than 100° was not a negative predictor of reparability. Consistent with our findings, Oh et al [34], reported that pseudoparalysis was not associated with poorer results in patients with large or massive tear. Unlike active flexion, better active external rotation was associated with a complete repair in our sample. Flexion is affected by a number of muscles, where external rotation in neutral is more specific to infraspinatus and would be more sensitive to pathology in the posterior cuff muscles.

In our study, tendon quality, and tear size and shape affected reparability. The crescent-shaped had the best outcome and U-shaped tears were more associated with a partial repair. The U-shaped tears have a larger mediolateral component which makes the convergence of tear edges more challenging. It has been reported that preoperative musculotendinous junction medialization, tendon retraction, and muscle quality are all predictive of tendon healing postoperatively when using a single-row rotator cuff repair technique [35]. Our study adds to the literature in this area as it explores the relationship between reparability and other non-surgical factor such as disability and clinical features used for clinical decision making.

In summary, failure to achieve a complete rotator cuff repair is multifactorial. Patient-reported disability and pre-operative clinical findings in conjunction with MRI investigations are potentially the most informative information for the clinicians. Intra-operative findings such as tear size, mobility, and shape are also critical but cannot be used to communicate the results prior to surgery. Future prospective studies are needed to incorporate the results of pre-operative MRI, the degree of weakness of the infraspinatus into clinical decision making process.

Among demographics, clinical examination and patient-oriented outcomes, the pre-op scores of the ASES and external rotation in neutral position were predictive of having a partially reparable large or massive tear. Surgical variables associated with a partial repair were tendon quality and tear size and shape. Patients with partial repair had a statistically significant improvement in range of motion and strength and all disability scores over time. However, they had less flexion and abduction and reported a higher level of residual disability at 2 years post-operatively.