1 Introduction

The Critical Shoulder Angle (CSA) as described by Moor et al. is measured on a true anterior-posterior shoulder radiograph (Grashey view) and is defined as the angle formed by a line connecting the bony glenoid baseline and a line drawn from the inferior bony margin of the glenoid to the most lateral border of the acromion.¹ Herewith, the CSA combines measurement of the glenoid inclination as well as the lateral extension of the acromion.¹

Based on their results and their biomechanical theory Moor et al. proclaimed that a CSA < 30° predisposes to osteoarthritis (OA) as the altered, more horizontal main force vector of the deltoid muscle would increase the contact pressure of the humeral head into the glenoid. Accordingly, Gerber et al. proclaimed a CSA > 35° to increase the risk for rotator cuff tears (RCT) since a longer acromion can give the deltoid muscle a force vector antagonizing the main force vector of the supraspinatus muscle.^1–3 Following the concept of the CSA, the edge of the acromion is more lateral in relation to the glenoid baseline. However, up to know it has not been further analyzed, whether with the more lateral extension of the acromion the position of the acromioclavicular joint (ACJ) is also changed. We therefore conducted a matched-pair analysis comparing the CSA and the horizontal position of the ACJ in patients with or without a full thickness tear of the supraspinatus tendon (RCT). The primary hypothesis was, that with a larger CSA not the acromion is disproportionally extended but the ACJ is positioned more lateral in relation to the glenoid.

2 Materials and methods

451 consecutive patients after arthroscopic shoulder surgery were screened for this retrospective matched-pair study.

2.4

2.4 Radiographic measurements

The radiographic analysis was conducted by two independent assessors, an experienced orthopedic registrar and an orthopedic consultant specialized in shoulder surgery who were blinded for the diagnosis of the patients. All preoperative true anterior-posterior radiographs were analyzed according to the Suter-Henninger classification and those not suitable for CSA measurement were excluded, respectively.⁵ The measurements were conducted using a certified radiographic image viewer (Sectra Workstation IDS7, Linköping, Sweden). The built-in angle and distance measurement features were used for the analysis. The CSA was measured as originally described by Moor (Fig. 1).¹ In addition, the distances from the glenoid baseline to the distal edge of the clavicle and the lateral edge of the acromion where measured in order to evaluate the horizontal position of the ACJ. Following the concept of the CSA, the angle between a line connecting the glenoid baseline and a line drawn from the inferior bony margin of the glenoid to the most distal edge of the clavicle was determined (distal clavicle angle; DCA) (Fig. 1).

Close

Fig. 1 True-ap radiograph of a right shoulder. The distances from the glenoid baseline to the lateral acromion (A) and the distal clavicle (B) were measured. The Critical Shoulder Angle (#) and a Distal Clavicle Angle (*) analyzed.

Export to PPT

3 Results

Of the total 451 consecutive patients, 162 (35.9%) were excluded with malpositioned preoperative radiographs. 139 (30.8%) patients could not be pair matched. Based on the above stated inclusion and matching criteria 150 patients (mean age at surgery 60.0 ± 9.5 years, range 45–80, 76 female (50.7%), male 74 (49.3%) could be included. Of those 150 patients, 75 with a mean age at surgery of 59.3 ± 7.9 years (range 45–80), had an arthroscopically validated full thickness RCT and underwent arthroscopic RCR. The control group with an arthroscopically validated intact rotator cuff consisted of 75 matched patient pairs with mean age at surgery of 59.4 ± 7.9 years (range 45–80) (Fig. 2).

Close

Fig. 2 Flow diagram of patient recruiting and matching.

Export to PPT

3.1

3.1 Position of the acromioclacivular joint (ACJ)

In all 150 patients, the mean distance from the glenoid baseline to the distal clavicle and therewith beginning of the ACJ was 14.6 ± 4.7 mm (range 1.2–28.5 mm) and the mean distance to the lateral acromion was 34.8 ± 4.7 mm (range 24.4–48.7) resulting in a mean, here called distal clavicle angle (DCA) of 13.6 ± 4.3° (range 1.3–24.7). In the RCT group, the distance to the distal clavicle was 15.4 ± 5.1 mm (range 1.2–28.5 mm), in the control group 13.7 ± 4.6 mm (range 2.3–22.2 mm, p = 0.034) resulting in an DCA of 15.1 ± 4.1° (range 5.2–24.7°) for the RCT and 12.7 ± 3.8° (range 2.6–20.1°, p = 0.0016) for the control group (Fig. 3). In patients with a CSA >35° (n = 57; 38.0%) the mean distance to the distal clavicle was 16.2 ± 2.9 mm (range 11.0–28.5) and 12.2 ± 4.0 mm (range 3.1–19.5) in the patients with a CSA < 30° (p = 0.002) (Table 1). This resulted in a DCA of 16.7 ± 3.6° (range 11.0–28.5°) in the patients with a CSA > 35° compared to 10.1 ± 4.0° (range 2.6–18.9°) in those with a CSA < 30° (p < 0.0001). The interrater reliability of the CSA and DCA measurements was calculated with 0.84.

Close

Fig. 3 Boxplot analysis of the Critical Shoulder Angle (CSA) and the Distal Clavicle Angle (DCA) and comparison between the Rotator-Cuff-Tear Group (RCT) and the control group. Horizontal line = median; ** = p < 0.01.

Export to PPT

Table 1 Overview of demographic data and results.

	RCT group (n = 75)	Control group (n = 75)	P
Patient age	59.3 years (45–80)	59.4 years (45–80)
Male gender	37 (49.3%)	37 (49.3%)
Distance to distal clavicle	15.4 ± 5.1 mm (1.2–28.5)	13.7 ± 4.6 mm (2.3–22.2)	0.034
Distal Clavicle Angle	15.1 ± 4.1° (5.2–24.7)	12.7 ± 3.8° (2.6–20.1)	0.0016
Critical Shoulder Angle	34.7 ± 4.3° (23.7–43.2)	32.8 ± 4.3° (23.2–41.5)	0.0018

4 Discussion

Up to now, the position of the ACJ and its relation to the CSA in patients with RCT has not been evaluated. Following the concept of the CSA, the present matched-pair study confirms a significantly larger CSA in patients with RCT as well as a significantly higher incidence of RCT in the cohort of patients with a CSA > 35° which was defined as a pathological cut-off by Moor.¹ However, as most important finding, the results reveal that the ACJ as well is positioned significantly more lateral in patients with RCT in comparison to the cohort with an intact rotator-cuff. Confirming our primary hypothesis, with a larger acromial cover reflected by the CSA, the ACJ is likewise positioned more lateral in relation to the glenoid and the acromion itself is not disproportionally enlarged. A more lateral position of the ACJ with subsequent further narrowing of the subacromial space might therefore be, among the CSA, an additional component in the multifactorial pathogenesis of degenerative RCT. For the ACJ it is well known that distally pointing osteophytes can lead to impairment of the supraspinatus tendon and eventually predispose to RCT.^6–9 Cuomo et al. found those ACJ osteophytes in over 60% of shoulders with full thickness RCT in comparison of age matched controls.⁷ As limitations we must consider that only anterior-posterior radiographs were analyzed and the detailed ACJ morphology with regard to osteoarthritis of the ACJ and related osteophytes was not evaluated. Herewith, any deduction on the clinical impact of the measured ACJ lateralization remains speculative at this point. Patients were matched according to gender, age and the presence of a reparable full thickness tear of the supraspinatus tendon. A more detailed matching according to tear size, retraction and fatty degeneration was not feasible due to the limited number of available patients in this single center study. Nevertheless, this study has several strengths: a strict matched-pair design to eliminate the bias of age and gender differences which is crucial when evaluating RCT in cohorts.

Patients with radiographs not suitable for CSA measurement according to the Suter-Henninger criteria were excluded and furthermore, repeated measurements by two independent blinded assessors were conducted. The diagnosis of RCT was based on the intraoperative evaluation with necessity for RCR and not only the preoperative MRI. In regard to the CSA, several studies have supported the findings by Moor that an increased CSA >35° is associated with RCT.^3,10–13 Garcia et al. reported that a larger CSA bears a higher risk for retear after RCR.¹⁴ Despite that, the concept and especially clinical relevance of the CSA is being critically discussed. Chalmers et al. found the mean CSA to be 34° in patients with RCT, compared to 32° in patients with intact rotator cuffs and concluded even though significant, that the difference was small enough that it could rather be influenced by measurement error in practice.¹⁵ Comparable, Bjarnison et al. report a non-significant difference in their CSA measurements with a mean of 33.9° in the RCT group and 33.6° in the control group.¹⁶ Shinagawa et al. analyzed the CSA in the Japanese population and hence, confirming a larger CSA as independent risk factor for RCT they found only small differences in their absolute values with 33.9 for the RCT versus 32.3 in their control group.¹⁷ Even though not associated to the ACJ, Kim et al. recently evaluated the combined influence of osteophytes protruding inferiorly from the undersurface of the anterolateral acromion and the CSA and on the incidence of RCT. Here, the risk for RCT was influenced more by the presence of those osteophytes than the CSA when both were evaluated together as related factors for RCT.¹³ Even though the present study confirms a significantly larger CSA in patients with RCT as well as a significantly higher incidence of RCT in the cohort of patients with a CSA > 35° the small differences in the absolute values leave the utmost clinical relevance in surgical decision making questionable. In regard to the relative position of the ACJ to the glenoid baseline, further studies should evaluate the influence of a more lateral positioned ACJ on the supraspinatus tendon which might point out to be an additional component in the multifactorial pathogenesis of RCT.

5 Conclusions

With an increase in the CSA, the ACJ is positioned more lateral in relation to the glenoid which might be an additional component in the multifactorial pathogenesis of RCT.

Ethics approval

There was a positive vote from the ethics committee of the Medical Faculty (6204R).

Funding

Not applicable.

Authors’ contributions

All authors were major contributors concerning the management of the patients, review of the articles and manuscript preparation. All authors read and approved the final manuscript.

CRediT authorship contribution statement

Hannes Kubo: Investigation, Formal analysis, Validation. Fariha Piela: Investigation, Formal analysis. Thilo Patzer: Writing - review & editing. Markus Konieczny: Validation, Visualization. Erik Schiffner: Writing - review & editing. Pascal Jungbluth: Writing - review & editing, Supervision. Rüdiger Krauspe: Supervision. Martin Hufeland: Writing - original draft, Conceptualization.