Concomitant Glenohumeral Pathologies in Patients with Acromioclavicular Joint Dislocations: How Do Acute and Chronic Instabilities Differ?

Background: Concomitant glenohumeral pathologies may be present in patients with acromioclavicular joint (ACJ) dislocations. This study aims to record and compare the prevalence and treatment of CGP in cases with acute and chronic ACJ dislocations. Methods: This retrospective cross-sectional binational, bicentric study included patients that underwent arthroscopically assisted stabilization for acute (group A) and chronic (group C) ACJ dislocations. Intraoperatively, CGPs and eventual treatments (debridement and reconstructive measures) were recorded. Results: The study included 540 patients (87% men; mean age 39.4 years), with 410 (75.9%) patients in group A and 130 (24.1%) in group C. Patients in group C were older (p < 0.001). The CGP prevalence was 30.7%, without a difference between groups A and C (p = 0.19). Supraspinatus tendon (SSP) and labral lesions were most common. Within group C, CGPs were more prevalent in surgery-naïve patients (p = 0.002). Among 49 patients with previous surgical treatment, CGPs tended to be more common in patients with prior open surgery than arthroscopically assisted surgery (p = 0.392). Increased CGP prevalence was associated with higher age (r = 0.97; p = 0.004) (up to 63% in the oldest age group, but also 17% for youngest age group) and higher in cases with Rockwood type-IIIB injuries compared to type-V injuries (p = 0.028), but type-IIIB injuries included more group C cases (p < 0.001). The most frequently found CGPs were treated by debridement rather than reconstructive interventions (SSP and labrum: p < 0.001, respectively). Conclusions: This study shows that one in three patients with ACJ instabilities has a CGP, especially elderly patients. Most of the CGPs were treated by debridement rather than constructive interventions.

After previous studies on CGPs in acute ACJ dislocations, a recent study suggested that chronic cases were associated with a higher prevalence of CGP, occurring in up to 53% of cases [26].Regardless of this finding, the (type of) treatment for the CGP (debridement vs. reconstruction) has been poorly described before.
The aim of our study was to describe and compare the prevalence of CGPs in patients with acute and chronic ACJ dislocations (RW types II-V) undergoing arthroscopically assisted ACJ stabilization and to investigate the treatment for CGPs.
In this retrospective, cross-sectional, binational and bicentric study, all patients who underwent arthroscopically assisted stabilization for acute and chronic ACJ instabilities (RW types II-V) between February 2007 and April 2019 were included consecutively.
We excluded patients with incomplete data and prior fractures, arthroplasty or tendon transfers of the shoulder girdle.
ACJ dislocations were classified radiographically [2,4].Differentiation between RW types IIIA/IIIB was also performed radiographically according to the presence (IIIB) or absence (IIIA) of an overriding of the clavicle on the acromion [4].
Patient parameters (age at time of surgery, gender, injury-to-surgery-interval, RW type, and previous surgery with respective type being open/arthroscopically assisted) were extracted from the particular clinic patient information system and coded into a centralized project database.
According to the injury-to-surgery-interval, patients were divided into two groups: acute (group A; 0-21 days) and chronic ACJ dislocations (group C; >21 days) [9][10][11][12][13][14]. Within group C, patients were divided into one group with previous surgical treatment for the ACJ dislocations and one without prior surgical treatment.Cases with previous surgical treatment were further differentiated by the respective type of prior interventions being open or arthroscopically assisted.According to patient age, several groups were formed (18-30; 31-40; 41-50; 51-60 and 61-71 years).
During surgery, a diagnostic glenohumeral arthroscopy was performed via a standard posterior viewing portal prior to the ACJ reconstruction in all cases.Relevant glenohumeral structures for evaluation included the rotator cuff (RC) [8,9,35,36] (Fox and Romeo classification [35]: type I: partial thickness tear; type II: complete tear of the upper 25 % tendon portion; type III: complete tear of the upper 50 % tendon portion; type IV: complete rupture of the tendon portion.Snyder classification of SSP lesions [36]: A: articular-sided tear; B: bursal-sided tear; C: complete tear), the labrum (Kim's lesion: incomplete or concealed posterosuperior labrum lesion [37]; glenolabral articular disruption, GLAD [38]), the long head of the bicep tendon (LHB) and its anchor (superior labrum anterior and posterior, SLAP, according to Snyder [39]: type I: frayed and degenerated superior labrum; type II: superior labrum and bicep tendon detached from glenoid rim; type III: SLAP lesion with a bucket-handle tear of the superior labrum, while the remaining labrum remains attached to the glenoid rim; type IV: SLAP lesion with an extension of the bucket-handle labrum tear into the biceps tendon), the pulley system (Habermeyer classification [40]: type I: isolated lesion of the superior glenohumeral ligament, SGHL; type II: lesion of the SGHL and a partial, articular-sided SSP lesion; type III: lesion of the SGHL and a partial, articular-sided subscapularis tendon lesion; type IV: lesion of the SGHL with a partial articular-sided lesion of the SSP and subscapularis) and the glenohumeral joint surfaces.Subacromial arthroscopy was only performed if anamnesis (also regarding symptoms prior to accident, e.g., impingement) or clinical findings (e.g., impingement/rotator cuff test) were suggestive for such.
These anatomic structures were routinely screened for possible CGPs and, if deemed relevant by the attending surgeon (for example, and trending towards high-grade lesions of the rotator cuff or labrum), treated during the same surgery.All findings and interventions were documented in the surgery reports.Reports were screened for CGPs and respective treatments, which were generally classified as debridement or reconstruction.If available, arthroscopic images were additionally evaluated by two independent examiners (M.M., M.S.) to double-check for pathologies.The character of the CGP was classified as either acute/traumatic or chronic/degenerative [12], to the extent possible.

Standard Analysis
Statistical analysis was executed using R software (v4.0.3).All study parameters were described using standard statistics including range and ratios, where appropriate.Age is reported with standard deviation.Following the testing of normal distribution (Kolmogorov-Smirnov test), group comparisons were performed by use of the Mann-Whitney-U test.The Chi-Square test (or Fisher's exact test, depending on sample size) was utilized to compare prevalences between groups, and, if appropriate, to compare the odds ratio (OR).For correlation analysis, the Pearson coefficient or the Spearman coefficient were calculated according to scale level (continuous vs. ordinal).Results were considered significant if p < 0.05.

Patient Cohort
A total of 540 patients (87% men; mean age 39.4 ± 12.1 years) were included in the study, comprising 410 (75.9%) patients in group A and 130 (24.1%) in group C (Figure 1).Patients in group A underwent surgery after a mean injury-to-surgeryinterval of 9 ± 4 days (range, 1-21 days) while the patients in group C underwent surgery after a mean injury-to-surgery-interval of 277 ± 201 days (range, 22-874 days).Patients in group C were older than patients in group A (43.8 ± 12.6 years compared to 38.0 ± 11.7 years; p < 0.001).
Regarding the two main groups according to RW type (IIIB; V), CGPs were more common in patients with a type-IIIB injury (37.4%; n = 61) as opposed to a type-V injury (27.8%; n = 99; p = 0.028; OR 1.51) at a comparable age (IIIB: 38.7 ± 12.8 years vs. V: 39.6 ± 11.9 years; p = 0.320).However, the RW type-IIIB group included a higher rate of group C cases [

Treatment of Concomitant Glenohumeral Pathologies
More than two thirds (69.2%; n = 115) of all cases with CGP were deemed relevant for treatment (Table 3).Treatment was performed more frequently in group A than in group C (p = 0.003; OR 2.88).In group A, SSPs were treated more frequently (p < 0.001; OR 18), without any difference in the rate of reconstructive measures or debridement (p = 0.392).
Lesions of the RC (SSP/SSC) and labrum (mainly SLAP lesions) were the most common type of CGP and were treated in 60 to 78% of cases when detected (Table 4).SLAP lesions were treated in 82.5% of cases (52/63).
For SSP and labrum lesions, debridement was carried out more often than reconstructive interventions (SSP: p < 0.001; labrum: p < 0.001).There was a similar tendency for SSC lesions (p = 0.166).

Discussion
The aim of our study was to describe and contrast the prevalence and therapy of CGPs in patients with acute and chronic ACJ instabilities.
Our study, featuring the largest cohort to our knowledge, demonstrated that approximately one in three patients with acute or chronic ACJ instabilities exhibits at least one CGP.Almost half of the patients presenting with a CGP were diagnosed with more than one pathology.Although most CGPs were classified as chronic/degenerative rather than acute/traumatic, most pathologies remained undetermined in their relation to trauma.Most of the CGPs were treated by debridement rather than reconstructive interventions.
CGP prevalence was lower in patients with prior surgery.There was a trend toward lower rates following arthroscopically assisted treatment, although it did not reach statistical significance.
The prevalences of CGP were higher in elderly patients.Patients with RW type-IIIB injuries had a higher rate of CGPs compared to type V but were confounded by a higher percentage of group C cases and a long injury-to-surgery-interval in the RW type-IIIB group.
Most CGPs, particularly in older patients, were identified intraoperatively by the surgeon, and 31% of all CGPs necessitated additional reconstructive surgical treatment.
Lesions of the SSP/SSC and labrum were most common and mainly treated by debridement rather than reconstructive measures.
Arrigoni et al. [25] were the first group to compare the prevalence of CGPs in acute and chronic cases, demonstrating no difference between groups.Jensen et al. [26] extended the analysis for acute and chronic cases (cut-off for time-to-surgery interval: 21 days) and results revealed a higher prevalence of CGPs for chronic cases.Both studies documented a higher incidence of CGPs in older patients, utilizing a cut-off age of 45 years established by Arrigoni et al. [25].
The current study corroborates the high rate of CGPs (albeit mainly low-grade) in surgically treated ACJ dislocations, although no difference between groups A and C was found.Age was confirmed as a significant factor linked to the presence of (possibly asymptomatic) CGPs [41].These CGPs can be eligible for surgical treatment in light of pain or functional reasons [42], including the consideration of eventual tear progression that can negatively affect these two aspects [43].However, the CGPs might also be treated non-operatively initially and re-evaluated later if they would present as symptomatic in the further course (in elderly patients).
We demonstrated no higher CGP prevalence in group C, in contrast to findings reported by Jensen et al. [26].However, Jensen et al. excluded patients with previous surgery while our group C included cases with previous shoulder surgery except for fractures, arthroplasty or tendon transfers.
Patients in group C with prior surgical treatment (37.7%) displayed a lower rate of CGPs than surgically naïve patients.Although it is presumed that CGPs were addressed during index surgery, validation would specifically apply to debridement in cases with previous arthroscopically assisted treatment.The therapeutic value such as rest, analgesia, physiotherapy or biomechanically improving surgery on the healing/subsequent absence of CGPs cannot be quantified.
Regarding RW types, Tischer et al. [29] and Jensen et al. [26] found tendencies for CGP to occur more often in types V.In our study, we found that type IIIB was linked to a higher CGP prevalence than type V cases at a similar age.However, this group exhibited a higher rate of group C cases and a longer injury-to-surgery-interval and a higher rate of SSP lesions.Missing differences in the sub-analysis of CGP prevalence within group A and C may be subject to a type-II-error.
These observations may be interpreted as an indicative of selection bias, considering that patients in the RW type-IIIB group primarily underwent non-operative treatment in the first place.These patients may have sought re-consultation later on in the chronic setting due to persistent complaints.
Although various aspects of residual symptoms, especially after non-operative treatments of ACJ dislocations, have been previously described [8,[30][31][32][33][34], the etiologic cause for (such) re-presentation after ACJ dislocations in our study is unknown and the cross-sectional design precludes an explanation.Our study implied a potential association of residual symptoms and CGPs, as 31.3% of CGPs were subject to further intraoperative reconstructive treatment.Reconstructive measures were more frequently undertaken in in patients with chronic than acute ACJ injuries.This percentage is comparable to Jensen et al. [26].
It is unclear if CGPs were caused by the same incident as the ACJ dislocation at a similar mechanism [39,[44][45][46], if they pre-existed and were possibly aggravated by the trauma or if they developed secondarily as a result of altered glenohumeral movement patterns [47][48][49][50].The uncertainty is reflected in the fact that 61.5% of all CGPs could not be classified as chronic/degenerative or acute/traumatic.
The use of arthroscopically assisted techniques allows simultaneous detection and treatment of CGPs.However, the value the of CGP treated (in the same session) is yet to be determined.Besides reconstructive procedures, especially the usefulness of less interfering treatment (debridement) remains largely unknown and should be subject to further research.The observation that the most common CGPs were addressed by debridement rather than reconstructive measures implied additional significance.
Several limitations should be considered before interpreting our results.The retrospective design introduces a potential for selection bias in this study, as CGPs might have led to symptoms and surgical intervention in a chronic setting.To reduce general selection bias, patients were extracted consecutively.The retrospective and descriptive study nature prevents making any implications about the etiology of CGPs or about treatment outcomes.In spite of the overall large sample number, a beta-error is assumed for the small group size of previously operated cases in group C. Subacromial arthroscopy was only performed in suspicious cases.
Despite these limitations, we were able to offer a comprehensive and large-scale study on the prevalence and type of treatment of CGPs in arthroscopically assisted ACJ stabilization.

Conclusions
This study, featuring the largest sample size to date to our knowledge, shows that one in three patients with acute and chronic ACJ instabilities has a CGP.CGPs were particularly prevalent in elderly patients, reaching almost two-thirds of cases.
Most of the CGPs were treated by debridement rather than constructive interventions.

Figure 1 .
Figure 1.Flowchart of the study cohort according to injury-to-surgery-interval and, for group C, previous treatment type.A, acute.ACJ, acromioclavicular joint.C, chronic.

Figure 1 .
Figure 1.Flowchart of the study cohort according to injury-to-surgery-interval and, for group C, previous treatment type.A, acute.ACJ, acromioclavicular joint.C, chronic.

Table 1 .
Prevalence and subclassification of concomitant glenohumeral pathologies in the overall group and for group A and C. A, acute; C, chronic; CGPs, concomitant glenohumeral pathologies; SLAP, superior labrum anterior and posterior; GLAD, glenolabral articular disruption.Bold p-values indicate significant results (p < 0.05).

Table 2 .
Percentage of patients with concomitant glenohumeral pathologies in different age groups.CGPs, concomitant glenohumeral pathologies.Bold p-values indicate significant differences (p < 0.05).