1 Introduction

Symptomatic rotator cuff disease is common, and while the underlying rationale of rotator cuff surgical repair when indicated is to improve functional patient outcome and restore structural integrity, 10–30% of rotator cuff tears (RCTs) are massive and irreparable.¹ Surgical options for massive RCTs may include an attempt at primary repair, though the reported failure rate in these cases is particularly high.² Reverse shoulder arthroplasty has demonstrated good outcomes in arthritic patients with massive RCTs, though it may not be the best option for younger, active patients.³ Superior capsular reconstruction (SCR), lower trapezius transfer and anatomical joint arthroplasty are other options.^4–7 Recently, several systematic reviews have demonstrated the clinical effectiveness of SCR as a surgical solution for irreparable RCTs,^8–11 though a review by Altintis et al.¹¹ outlined the relatively poor quality of many of the existing studies and notable complications that have been reported.

First described by Gerber et al., in 1988,¹² latissimus dorsi tendon transfer (LDTT) has become an established treatment option for patients with massive and irreparable RCTs.^13–27 Factors such as pre-operative shoulder function and general strength may influence outcome.²⁸ However, given the transferred latissimus will provide an inferior force that must be counterbalanced by the deltoid and subscapularis,¹² an intact subscapularis at the time of surgery has been considered imperative.²⁹ The inferior outcomes reported in patients undergoing LDTT with concomitant subscapularis insufficiency have been reported.^{15,18,19,22,23} However, it is important to note that the finding of worse outcome with subscapularis deficiency in reported studies appears largely based upon analysis of outcomes. While studies published by Gerber et al.^22,23 in 44 and 67 patients reported inferior outcomes in patients with a deficient subscapularis, other reports were only small studies by Weening et al.,¹⁵ (16 patients, of which 3 of 4 patients with subscapularis deficiency had unfavourable results) Aoki et al.¹⁸ (12 patients, 1 of 2 patients with subscapularis deficiency had a poor outcome) and DeCasas et al.¹⁹ (14 patients, of which the only patient with subscapularis deficiency had a poor outcome).

Given the relative lack of research specifically evaluating post-operative LDTT patient outcome associated with subscapularis insufficiency, the current study sought to investigate clinical outcome and patient satisfaction at a minimum 2-years post-surgery, in patients undergoing LDTT for massive irreparable RCTs with varied subscapularis integrity. It was hypothesized that LDTT would be just as clinically effective for patients with subscapularis insufficiency (partial or full thickness tearing), as to those with an intact subscapularis.

2 Methods

2.1

2.1 Patients

Between 2004 and 2018, a total of 118 patients presenting with massive RCTs that had been deemed irreparable underwent primary LDTT with a single surgeon (PC). Of the 118 patients that underwent surgery over the nominated period, patients were subsequently contacted for this retrospective review until the required sample size (n = 22 per group, based on an intact or deficient subscapularis as described below) was attained. Fig. 1 displays the flowchart of patient contact and assessment. Therefore, a total of 48 patients were included in the current study, including 22 patients with an intact subscapularis at the time of LDTT and 26 patients with subscapularis insufficiency (Fig. 1 and Table 1) as determined below.

Close

Fig. 1 Study flow chart demonstrating patient recruitment and assessment.

Export to PPT

Table 1 Patient demographics and clinical scores based on subscapularis grouping (intact or deficient). Shown are means and p-values.

Variable	Group 1 (Intact, n = 22)	Group 2 (Deficient, n = 26)	p value
Gender, males/females n (%)	15/7 (68.2/31.8)	19/7 (73.1/26.9)	N/A
Dominant Limb is Operated Limb, n (%)	13 (59.1)	15 (57.7)	N/A
Age at surgery (years), mean (range)	56.9 (45–68)	57.4 (43–75)	0.617
Post-operative Review Time (months), mean (range)	79.6 (24–150)	69.9 (24–156)	0.555
Concomitant Surgery, n (%)
Subacromial Decompression	8 (36.4)	8 (30.8)	N/A
Long Head of Biceps Tenotomy	2 (9.1)	4 (15.3)	N/A
Acromioclavicular Joint Excision	7 (31.8)	6 (23.1)	N/A
UEFI, mean (range)	69.6 (40–80)	67.0 (31–80)	0.265
GRC, mean (range)	3.6 (1–5)	3.1 (-2–5)	0.094

Massive and irreparable RCTs were confirmed pre-operatively in all 48 patients included in the current study, via Magnetic Resonance Imaging (MRI) alone (n = 36), MRI and ultrasound (US) (n = 7), US alone (n = 4) and computed tomography (CT) arthrogram (n = 1). All imaging was further categorised based on whether subscapularis was intact or insufficient. For the current study, an intact subscapularis was defined as ‘intact’ (n = 22), while a deficient subscapularis was defined as pre-operative imaging demonstrating and reporting a ‘partial tear’ (n = 16) or a ‘full thickness tear’ (n = 10). This research was approved by the relevant Institutional Review Board (IRB).

3 Results

No significant differences (p > 0.05) were observed in patient age at surgery or the time of post-operative review between the two surgical groups, with a similar distribution of concomitant procedures between the two groups (Table 1). Furthermore, no significant group differences were observed in either the UEFI or GRC (Table 1). High levels of patient satisfaction were observed in both groups, across all satisfaction items (Table 2).

4 Discussion

The current study has demonstrated encouraging post-operative clinical outcomes and high levels of patient satisfaction after LDTT for massive and irreparable RCTs, similar to existing studies.^13–27 However, while prior studies have reported inferior outcomes in the presence of subscapularis insufficiency,^{15,18,19,22,23} the most important finding from this study is that no significant difference in clinical outcomes could be demonstrated between those undergoing LDTT with, or without, subscapularis insufficiency at the time of surgery.

Gerber et al.¹² first reported the use of LDTT in 1988, with subsequent studies in 69²³ and 46²² shoulders at a mean 53 and 147 months, respectively, demonstrating good outcomes as reported by the Constant Score and Subjective Shoulder Value (SSV). In other large reported cohorts, Kany et al.¹³ reviewed 62 patients and reported good outcomes reported by the Constant, SSV, Simple Shoulder Test (SST), American Shoulder and Elbow Surgeons (ASES) score and a Visual Analogue Pain Scale (VAS), with 81% of patients satisfied at a minimum 2 year follow up. El-Azab et al.²¹ investigated 93 shoulders at mean 9.3 years with improved Constant, ASES and VAS scores. While El-Azab et al.²¹ reported better outcomes in younger patients, Kany et al.¹³ specifically compared outcomes in patients ≤55 and ≥ 75 years of age, demonstrating no group differences. Smaller studies undertaken by Postacchini et al.²⁵ (7 patients), Aoki et al.¹⁸ (12 shoulders), Clavert et al.¹⁴ (14 patients), DeCasas et al.¹⁹ (14 patients), Weening et al.¹⁵ (16 patients), Miniaci et al.²⁴ (17 patients) and DeBeer et al.²⁰ (26 shoulders) all demonstrated clinical improvement at various post-operative time-points resulting from LDTT reported via various outcome measures, including the Constant score, University of California at Los Angeles (UCLA) shoulder score, SSV, ASES and various satisfaction scores. Valenti et al.²⁶ reviewed a mix of primary (17 patients, 85% satisfied) and revision (8 patients, 50% satisfied) LDTT cases with mixed outcomes based on prior surgery, while Birmingham et al.¹⁷ performed LDTT as a salvage procedure in 18 patients after failed rotator cuff repair, reporting significant improvement in pain and active arm elevation. The current study demonstrated high rates of patient satisfaction, though no other study employed the UEFI or a GRC scale making comparison in clinical outcome to other studies difficult. At the time of development of the current study, given the vast array and inconsistency of PROMs already employed to assess patients after LDTT, we chose the UEFI as a well-reported score for upper limb function as a supplement to the GRC and satisfaction rating scores employed.

Again, specific to the current study we were unable to demonstrate a significant difference between patients with, or without, an intact subscapularis at the time of LDTT. In a review published by Grimberg et al.,³³ it was noted that studies reported various ways of dealing with the presence of concomitant subscapularis tears, including complete patient exclusion from LDTT, or repairing small tears of the upper third subscapularis. El-Azab et al.²¹ investigated 93 shoulders at mean 9.3 years, of which 13 patients underwent concomitant repair of a partial subscapularis lesion, with no clinical difference seen between these patients and those that had an intact subscapularis at the time of surgery. Studies by Gerber et al.^22,23 reported inferior outcomes in the presence of subscapularis insufficiency, also stating that if subscapularis function is deficient, the procedure is of questionable benefit and probably should not be used.²³ As previously mentioned, other studies reporting inferior outcomes with a deficient subscapularis are in small patient cohorts,^15,18,19 and it should be further highlighted that across all of these studies the number of patients undergoing LDTT with a deficient subscapularis, relative to the total patient group, was very low. Nonetheless, a more recent study published by Mihata et al.³⁴ in a large cohort of patients that had undergone SCR reported that the presence of subscapularis tears negatively affected clinical outcomes (including PROMs, strength and range of motion) and complication rates. Therefore, larger studies with more robust objective outcome measures and in the presence of varied surgical treatment options may be required.

The biomechanical role of subscapularis in LDTT has been previously discussed.³⁵ In a biomechanical cadaveric model, Werner et al.³⁵ reported that the subscapularis had a significant impact on glenohumeral translation and humeral head rotation, suggesting that the inferior results of LDTT in the presence of subscapularis dysfunction may be explained by the loss of humeral head stability upon arm abduction and elevation. The translation of these findings specifically to an in vivo model are hard to determine. Regardless, the reduced stabilizing role otherwise provided by an intact subscapularis did not appear to significantly affect patient outcome and post-operative satisfaction in the current study.

A number of limitations must be acknowledged in the current study. Firstly, the current study included 48 patients (with recruitment ceasing once the desired sample size for each group was attained). We appreciate this may present a selection bias, though it should also be acknowledged that of the initial 54 patients contacted and invited to participate, there were only 6 patients that had declined. Studies with larger patient cohorts exist^13,21,23 albeit the majority are in small patient cohorts, and larger studies with more comprehensive patient outcomes will benefit future research. Second, the current study sought to investigate the UEFI, GRC and patient satisfaction. Several other outcome measures have been employed (with the Constant score the most common) making comparison of clinical outcomes in the current study difficult to compare to existing literature. As discussed earlier, given the vast array and inconsistency of PROMs already employed to assess patients after LDTT, we chose the UEFI as a well-reported score for upper limb function as a supplement to the GRC and satisfaction rating scores employed. Further to this, the nature of this retrospective review (and in some patients that were 13 years post-surgery at the time of the review) made it difficult to evaluate all patients face-to-face, with many requiring a comprehensive phone review given the fact that many now lived in regional areas. This also limited our ability to undertake any objective measures such as strength and/or range of motion assessment. Other studies have also sought to review post-operative radiographic changes,^13,22 which the current study did not and may be of value to evaluate glenohumeral joint changes in the presence of the theorized reduction in stability not provided in the presence of subscapularis insufficiency. Again, similar to the lack of collected and reported objective clinical measures in the current retrospective study, radiological review was difficult and should be undertaken in future studies to better assess its effect on patient outcome.

5 Conclusion

The current study demonstrated that LDTT performed for massive irreparable rotator cuff tears resulted in encouraging clinical outcomes, in support of the existing literature. However, while the current study has limitations in that only the UEFI, GRC and Patient Satisfaction were assessed as previously discussed, it still demonstrated high levels of patient satisfaction in patients undergoing LDTT irrespective of the underlying level of subscapularis integrity. While smaller studies have reported less favourable outcomes in patients undergoing LDTT with subscapularis insufficiency,^{15,18,19,22,23} larger comparative studies with more robust and activity-specific assessment tools, together with serial radiographic imaging, may be required to truly evaluate the mid- and long-term stabilizing benefit offered by an intact subscapularis.

Ethical approval

This research was approved by the University of Western Australia (RA/4/20/4235) Human Research Ethics Committee (HREC).

Funding

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