Do Acromiohumeral Centre Edge Angle and Greater Tuberosity Angle Correlate With the Full-Thickness Degenerative Supraspinatus Tear?

Background: Many radiographic parameters associated with the extrinsic cause of supraspinatus tear have been proposed. The aim of this study was to correlate the relationship between full-thickness degenerative supraspinatus tear (FTDST) and the patient’s radiographic parameters, including the acromiohumeral centre edge angle (ACEA) and the greater tuberosity angle (GTA). Methods: A retrospective study was conducted. We included 116 patients who had undergone shoulder arthroscopic surgery at our institute. The case group included FTDST patients, while the control group also included patients without evidence of supraspinatus tear. In each patient, the ACEA and GTA values were measured and analysed by two independent observers. Intra-inter observer reliability was assessed. Multivariate regression analysis was performed. Results: The ACEA values were signicantly higher in FTDST, with a mean of 26.44°± 9.83° compared with 16.81° ± 7.72° in the control group (P < 0.001). Multivariate regression analysis also showed that higher ACEA values were associated with a FTDST (odds ratio 1.16 per degree, P = 0.01). Meanwhile, for GTA values, a statistically signicant difference was found with a mean of 70.92° ± 6.64 compared with 67.84° ± 5.56 in the control group (P = 0.02). However, Stepwise regression analysis rejected GTA as a predictor for FTDST. Conclusions: Our study demonstrated that the presence of higher ACEA values is an independent signicant risk factor for the presence of FTDST. Consequently, GTA values may be less helpful in assessing the risk of FTDST, especially in this specic population. population. Pearson used to identify whether any correlation was present between age and ACEA among this total population. We found a negligible correlation between ACEA value and age According to logistic regression analysis, an increased ACEA has an independent risk of FTDST, with an odds ratio of 1.13 per degree. In addition to higher ACEA values, the analysis that increased age was a risk factor in FTDST. Our results regarding patient age aligned with those of prior studies, which revealed that the prevalence of RCT correlated with patient age (18). The results of in the present study with FTDST are comparable with those in the study Singleton et reported a positive correlation between ACEA and acute traumatic RCT (23.9 vs. p < 0.001) not RCT size ACEA on the chronic degenerative tear still doubtful. our results, certied ACEA a FTDST used reliable measurement detecting FTDST on standard plain radiographs. between higher ACEA values and rotator cuff in the same way Acromial and critical shoulder (CSA). was explained by the vertical force vector of the middle bre of the deltoid muscle, where the pull, the lateral extension of the acromion, was directed upward, lead SSP tear compression a tear


Background
Rotator cuff tear (RCT) is a common cause of chronic shoulder pain and disability (1). Extrinsic causes for RCT are usually associated with subacromial impingement (2), de ned by the supraspinatus (SSP) tendon becoming entrapped between the acromion process and the greater tuberosity. As concern about SSP pathology increased, many previous studies focused on excessive lateral acromial coverage and con rmed that it is associated with a higher incidence of RCT (3)(4)(5). Regardless, the complex interlinkage of these acromial morphology parameters with RCT is still under extensive exploration. Recently, Singleton et al. (6) introduced the acromiohumeral centre edge angle (ACEA), a new measurement to be used with true AP shoulder radiography. The lateral projection of the acromion coverage humeral head, as a valid measurement with good reproducibility, has an accuracy comparable to measurement using a computed tomography scan. The results showed that the ACEA value was signi cantly higher in patients with RCT. However, because the Singleton et al. study only included acute traumatic RCT, the usefulness of ACEA in predicting degenerative RCT is still questionable. Furthermore, the subacromial impingement process necessarily comprises two sides of the bony structure, making evaluation of both bony sides, the acromion, and the greater tuberosity (GT) equally important. GT morphology still plays an important role in subacromial impingement. Some studies report that a displaced malunion GT was related to the worst outcome (7) and that the tuberosity procedure may provide a satisfactory result for irreparable massive rotator cuff tear (8). Additionally, the greater tuberosity angle (GTA), a new radiographic marker that evaluates the GT's position proposed by Cunningham et al., has been advocated as a reliable predictor marker for RCT. They suggested that the development of degenerative RCT (9) was associated with a higher GTA values. However, their study group included both participants with partial-thickness RCT and participants with full-thickness RCT. According to a recent study (10), some acromion parameters are associated with full-thickness but not with partial-thickness RCT. For this reason, the speci c full-thickness degenerative supraspinatus tear (FTDST) subgroup correlation with these parameters should be determined. Hence, no previous study has considered any of these values, including ACEA and GTA, as a speci c tool to be used with patients with FTDST. Moreover, the literature contains limited information about these parameters, especially those pertaining to a Southeast Asian population. Certain authors have attempted to evaluate the correlation of both parameters to the incidence of FTDST and analyse any association of both parameters with patients' demographic data and arthroscopic ndings. The primary objectives of our study were (1) to evaluate the presence of signi cant differences regarding the ACEA and the GTA values among patients with or without FTDST and (2) to assess the association between any of these parameters with other variables, such as age, sex, and SSP tear retraction. The hypothesis was that higher values for ACEA and GTA would be associated with a higher likelihood of detecting FTDST.

Materials And Methods
Sample A retrospective study was conducted in Ramathibodi Hospital, Mahidol University, Thailand. The medical records of all patients who had undergone shoulder arthroscopic surgery between April 2016 and July 2018 were collected. All pre-operative true AP shoulder radiographs were analysed. The patients selected were divided into two groups. The case group consisted of individuals presenting with a clinical diagnosis of FTDST, which was con rmed by history, preoperative-magnetic resonance imaging (MRI) scans, and intraoperative arthroscopic ndings. The control group consisted of those with shoulder pain without any nding of RCT based on history; physical examination; preoperative-MRI scans; and arthroscopic ndings, such as labral injury, shoulder instability, and primary adhesive capsulitis. Patient demographic data included age, gender, bicep pathology, fatty degeneration of the SSP by Goutallier staging (11), and SSP tear retraction in the frontal plane grading by Patte classi cation (12).

Inclusion criteria&exclusion criteria
The inclusion criteria were that the patients had a pre-operative true AP shoulder radiograph with the proximal humerus in an acceptable rotation of the affected shoulder. The exclusion criteria included partial-thickness RCT, any history of traumatic event to exclude any potential traumatic aetiology, previous surgery, fractures, and/or dislocation around the shoulder, congenital shoulder deformity, shoulder tumours, or infection. Patients with any evidence of osteoarthritis change in the glenohumeral joint were also excluded due to the possibility of producing outliers of these parameters.
Patients meeting the inclusion and exclusion criteria were recruited. Therefore, the nal sample comprised 116 patients who were enrolled into the study. The case group included 84 patients, and the control group included 32 patients. The studied population was Thai and exclusively Southeast Asian. This study was ethically approved by our hospital's institutional research board committee (IRB number MURA2018/837). All methods in the study were carried out in accordance with the Helsinki guidelines and relevant CIOMS guidelines.

Data collection & outcome measurement
ACEA and GTA measurements were determined on true AP shoulder radiographs and described, as shown in Fig. 1. The ACEA is de ned as the angle between a line drawn superiorly from the centre of the humeral head parallel to the glenoid and a line from the centre of the humeral head to the acromion's outer edge (Fig. 1B) (6). The GTA represents the angle between a line parallel to the humerus diaphysis passing through the centre of rotation of the humeral head and a line connecting the upper edge of the humeral head to the most superolateral edge of the GT (Fig. 1C) (9). All measurements were performed by two independent assessors: an orthopaedic surgeon specializing in the shoulder and a radiologist specializing in musculoskeletal imaging) using a goniometer tool in the Picture Archiving and Communication System (PACS). Both were blinded from intra-operative ndings. Then a repeated measurement was performed with one-month interval. Inter-observer and intra-observer reproducibility was determined.
In an effort to avoid the effect of the rotation of the proximal humerus, the parameters were measured on the true AP shoulder radiograph. Each patient was placed in a supine position with the arm adducted to the side and in a neutral position. Rotation in the axial plane was accepted up to ± 20 . This protocol was due to unchanged ACEA and GTA parameters in this rotation range according to previous studies (6,9).

Statistical analysis
Statistical analyses were calculated using Stata 16 software (StataCorp, College Station, TX, USA). The difference in the ACEA and GTA of the case and control groups was determined using independent t-tests. Correlations between parameters and age were assessed with the Pearson correlation coe cient method. Receiver operating characteristic (ROC) analysis curves were devised to determine the diagnostic ability of these parameters. Multivariate adjusted analysis by logistic regression analysis was used to determine each factor for the occurrence of FTDST with respective odds ratios. A Bonferroni post-hoc test was performed to evaluate if there was any signi cant difference among the subgroup analysis based on the SSP's tear retraction grading. The limits of agreement between two assessors were examined with Bland-Altman plot analysis. The intra-rater reliability and inter-rater reliability were assessed using the intraclass correlation coe cient (ICC). The ICC was interpreted as follows: 0 to 0.40, poor; 0.41 to 0.60, moderate; 0.61 to 0.80, good; and 0.81 to 1.00, excellent (13).

Patient demographic data
The total number of recruited samples was 116 shoulders. The case group included 84 shoulders representing shoulders with FTDST, while the control group (without evidence of SSP tear) included 32 shoulders. The mean age was 64.19 ± 7.67 years (range, 46-81 years) for the case group and 35.81 ± 14.13 years (range, 15-65 years) for the control group.
In every shoulder, we identi ed the presence of SSP tear, SSP tear retraction, and biceps pathology. In our series, 61.9% had biceps pathology in the case group compared to 12.5% in the control group. There were 42 patients with SSP retraction grade 1 (50%), 33 patients with SSP retraction grade 2 (39%), and 7 patients with SSP retraction grade 3 (8%). All patient characteristics are listed in Table 1.  For the assessment of any correlation between age and these parameters among total populations, a small positive strength of association (coe cient < 0.3) was found between age and both parameters. However, the results showed no statistical differences between these parameters within each group, as seen in Table 3.  The ROC curves were designed to evaluate the ability of both angles to predict FTDST.
The curves showed that an ACEA was a good predictor for FTDST with an area under curve as 0.78. For a GTA value, the area under the curve was 0.67, which is interpreted as a fair predictor for FTDST ( Figure 2). To determine the cut point, an ACEA value of 18° was a good predictor for full-thickness SSP tear, which had 85% sensitivity and 50% speci city, while a GTA value of 68° had 77% sensitivity and 44% speci city. This nding indicated that the ACEA value is a more accurate diagnostic test than is the GTA value. The differences in the cut-off values of ACEA and GTA are reported in Table 5. The multivariate analysis showed that the ACEA parameter was the only parameter that was found to be statistically signi cant. A higher ACEA value indicated an increased risk of FTDST with an odd ratio of 1.16 per degree (P = 0.01). However, there was no statistical signi cance for the GTA parameter (P = 0.10) ( Table 6). The risk factor for FTDST was increased age (odd ratio of 1.26 per year; p < 0.001). Our ndings also showed that while the mean ACEA and GTA values of the FTDST group were larger than those of the control group, the means of the parameters among subgroups categorized by Patte classi cation did not show a signi cant difference. A comparison of both parameters is shown in Figure 3. Reliability testing for the ACEA and GTA values showed that the mean ACEA difference and the mean GTA difference were -0.94 ± 3.2 and 0.60 ± 2.0, respectively. The ICC values for the ACEA and GTA measurements were 0.95 and 0.94, respectively (Table 7). Interobserver reproducibility between both assessors was excellent for corresponding to previous reports (6,9). The Bland-Altman plot of the mean difference between the repeated measurements is shown in Figure 4. Discussion RCT is one of the most common causes of chronic shoulder pain, leading to decreased functionality, declined quality of life and escalated utilization of health care resources (14). Due to its cost effectiveness and accessibility, the standard shoulder series is typically the rst line of investigation for the patients suspected RCT (15) to provide additional information and needs of advanced images as MRI of the shoulder (16). This study aims to correlate between the radiographic parameters, as ACEA and GTA, from the standard shoulder radiographs and FTDST in those patients who presented with shoulder pain underwent arthroscopic surgery.
Our results showed that the mean ACEA parameter was 26.44° ± 9.83° in the case group compared to 16.81° ± 7.72° in the control group, with a statistically signi cant difference between the groups. Female gender had a statistically signi cant higher ACEA value when comparing to male gender in the total population. Pearson correlation was used to identify whether any correlation was present between age and ACEA among this study's total population. We found a negligible correlation between ACEA value and age (17). According to logistic regression analysis, an increased ACEA has an independent risk of FTDST, with an odds ratio of 1.13 per degree. In addition to higher ACEA values, the analysis manifested that increased age was a risk factor in FTDST. Our results regarding patient age aligned with those of prior studies, which revealed that the prevalence of RCT positively correlated with patient age (18). The results of in the present study with FTDST are comparable with those in the study by Singleton et al., which reported a positive correlation between ACEA and acute traumatic RCT (23.9 vs. 16.6, p < 0.001) (6). However, they did not consider RCT size and mentioned that implementation of ACEA parameter on the chronic degenerative tear is still doubtful. With our results, we certi ed that ACEA could be generalized to a FTDST population and used as a reliable measurement tool for detecting FTDST on standard shoulder plain radiographs. The association between higher ACEA values and rotator cuff pathology could be clari ed in the same way that it has been for other parameters, such as the Acromial index (AI) and critical shoulder angle (CSA). This was explained by the vertical force vector of the middle bre of the deltoid muscle, where the pull, in uenced by the lateral extension of the acromion, was directed upward, lead plausibly to SSP impingement and consequence tear due to a compression effect causing a degeneration tear of the rotator cuff, as stated in previous literature (19,20).
Despite the high variability of GT morphology (21), the GTA parameter was rst proposed in 2018 for a new reliable radiographic marker of degenerative RCT. However, the precise relationship between GTA's high values and the incidence of RCT is not yet understood. In the present study, GTA was found signi cantly between groups with a small size difference; the mean in the case group was 70.92° ± 6.64, compared to 67.84° ± 5.56 in the control group.
However, multivariable analysis showed no statistically signi cant correlation between GTA and FTDST. Hence, stepwise multiple regression analysis rejected GTA as a predictor for FTDST. The result of the present study contrast with those reported by Cunningham et al., who noted that the mean GTA parameter was 72.5° ± 2.5° in the RCT group compared to 65.2° ± 4.1° in the control group and concluded that degenerative RCT in the European population was associated with GTA values of more than 70° (9). In addition, Yoo et al. reported that high GTA values accompanied RCT in the Korean population, based on MRI results (22). This discrepancy between results could be explained by the population-based anatomic variation that may exist in a Southeast Asian population, such as the Thai population. A recent study has documented that the Asian population exhibits a smaller humeral head compared to the Western population (23); these ndings could possibly explain why the GTA effect is larger in Asians than in Europeans (22). Moreover, previous studies examined the GTA effect on individuals with partial-or full-thickness supraspinatus tears, but our study examined the GTA effect only on FTDST; partial tears were not taken into consideration. Recent studies (5, 10) demonstrated a dissimilarity association between acromial parameter and type of RCT tear (full thickness vs. partial thickness). Additionally, Seo et al. (24) reported that the mean GTA values for bursal-side partial RCT tended to be larger than those for full-thickness RCT. Therefore, the different populations' characteristics could affect the magnitude of the parameter and may have produced different ndings among studies as well This study was the rst to evaluate these angles based on arthroscopic ndings for their usefulness in determining the risk of FTDST. The most important nding was that ACEA was shown to have a statistically signi cant, high association with FTDST. Stepwise logistic regression analysis demonstrated that between ACEA and GTA, only ACEA appears to be valid in correlation with FTDST: GTA was rejected as a factor in predicting FTDST. No previous study has performed logistic regression to assess the effects of these parameters on FTDST. We summarized that ACEA could be used as an independent factor to assess the risk of FTDST in our study. As such, we also concluded that GTA could not be utilized to assess the risk of FTDST in the Southeast Asian population, particularly in the Thai population. Nevertheless, a positive correlation was not found between a higher value and SSP tear retraction grading. Thus, these parameters cannot be used to differentiate the severity of tear retraction of SSP. Although the possibility of type II errors should be considered, further biomechanics study is essential for speci c assessment of the in uence of these parameters on tear retraction.
The present study had several limitations. First, due to limitation of retrospective study nature and arthroscopic based study design, some patient's demographic data might have a difference between the groups. FTDST patients tend to be females who are older than patients with labral injury and shoulder instability, even adhesive capsulitis, which is mainly included in the control group. Despite this variation, the statistically signi cant difference still endured for the ACEA parameter after multivariable analysis. Second, our study has relatively small sample size compared to previous studies. However, a statistical power analysis was performed for sample size estimation, with alpha = .05 and power = 0.80. The sample size needed with this effect size was approximately 25 samples within each group. These were su cient for detecting a difference of 5° between groups if the standard deviation of each group was 7.7., despite the fact that our proposed sample size was more than adequate for this study's primary objective. Third, the difference of ACEA and GTA between the patients with and without FTDST was approximately Page 11/14 9.6° and 3.1°, respectively; these results were within their SD range. Although the ndings were statistically signi cant, careful interpretation is essential to determining the clinical application of these parameters

Conclusion
The presence of higher ACEA values is a signi cant independent risk factor for the presence of full-thickness, degenerative SSP tear. The measurement of ACEA could be a useful tool to determine additional advanced images such as MRI to con rm diagnosis of this condition in a patient with chronic shoulder pain. Consequently, GTA values may be less helpful in assessing the risk of full-thickness, degenerative SSP tear, especially in the Southeast Asian population.

Consent for publication
Not applicable.
Availability of data and material The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.