Thickening of the axillary recess capsule on ultrasound correlates with magnetic resonance imaging signs of adhesive capsulitis

Abstract

Objective

To correlate the thickness of the axillary recess capsule measured by ultrasound with magnetic resonance imaging signs of adhesive capsulitis in patients with shoulder pain.

Materials and methods

We prospectively evaluated 193 consecutive patients (141 women and 52 men, aged 40–69 years) with shoulder pain lasting 1–9 months from January 2015 to December 2016 who underwent shoulder ultrasound. All participants had routine shoulder ultrasound with additional measurement of axillary recess capsule thickness. After examinations, two groups were formed: negative ultrasound group, composed of patients with a capsule thickness of 2.0 mm or less, and positive ultrasound group, composed of individuals with a capsule thickness greater than 2.0 mm. All patients from the positive ultrasound group and 27 randomly chosen patients from the negative ultrasound group underwent shoulder magnetic resonance imaging.

Results

In all, 169/193 patients (88%) had an axillary recess capsule thickness of 2.0 mm or less (negative ultrasound group) and 24/193 patients (12%) had a capsule thickness greater than 2.0 mm (positive ultrasound group). Twenty-seven patients from negative ultrasound group (27/169) were randomly selected to undergo shoulder magnetic resonance imaging. None of them had magnetic resonance imaging criteria for adhesive capsulitis. All patients from positive ultrasound group (24/24) underwent shoulder magnetic resonance imaging and 23 of them (23/24) had magnetic resonance imaging signs of adhesive capsulitis, with a sensitivity of 100% and a specificity of 96%.

Conclusion

In patients with shoulder pain, a thickness greater than 2.0 mm of the axillary recess capsule measured by ultrasound correlates to magnetic resonance imaging signs of adhesive capsulitis with good sensitivity and specificity.

Keywords

Ultrasound musculoskeletal diagnostic imaging adhesive capsulitis axillary recess capsule

Introduction

Adhesive capsulitis (AC) is an inflammatory and fibrosing condition of the shoulder characterized by pain and decreased range of movement (ROM),¹ with an incidence ranging from 2% to 5%.² AC begins with an inflammatory process of the glenohumeral synovial membrane that typically affects the joint capsule, the coracohumeral ligament (CHL), and the rotator cuff interval.³

The main symptoms of AC include persistent pain and gradual joint stiffness, with loss of both active and passive ROM, especially with limitation of external rotation of the arm.^4,5 Neviaser and Neviaser,⁶ and Neviaser and Hannafin⁷ divided AC into four clinical stages, which initiate with shoulder pain and progresses to movement restriction (Table 1). AC is considered a self-limiting condition; however, some studies have shown that symptom resolution may take up to two years and a residual restriction of movement may be seen in 50% of patients.⁸

Table 1.

Clinical stages of adhesive capsulitis described by Neviaser.⁶

Stage 1	Duration of symptoms: 0–3 months Pain during active and passive range of movement (ROM) Pathology: hypertrophic, hypervascular synovitis with a relatively normal joint capsule
Stage 2	Duration of symptoms: 3–9 months Chronic pain exacerbated by movement. Progressive worsening of motion restriction Pathology: synovium is hypervascular with scar formation in the underlying capsule
Stage 3	Duration of symptoms: 9–15 months Minimal pain. Significant limitation of ROM Pathology: capsule is thickened and fibrotic with articular volume reduced
Stage 4	Duration of symptoms: 15–24 months Minimal pain. Progressive improvement of ROM Pathology: capsular remodeling occurs over the following six months or more

Clinical criteria are traditionally considered the standard for diagnosing AC. However, in the early stages, AC can be clinically confused with other shoulder conditions, such as labral tears, rotator cuff disease or neuropathy (cervical or peripheral), in such cases, imaging modalities are especially useful.⁹

Magnetic resonance imaging (MRI) is considered an excellent diagnostic imaging modality. Typical findings of AC on MRI are thickening and abnormal signal intensity of the joint capsule in the axillary recess and rotator interval as well as thickening of the CHL.^4,9–13

Although several MRI findings for AC have been reported, the role of ultrasound (US) has not yet been established and only a few studies have tried to prove its accuracy, based on the evaluation of the thickening of the CHL and increased vascularity in the rotator interval.^14,15

Only two recent reports in literature approached the diagnosis of AC based on ultrasound changes in the axillary recess capsule,^16,17 even though MRI has shown that capsular thickening in this location is the most relevant imaging finding for AC.^13,18 The real relevance of US evaluation of the axillary recess for diagnosing AC is still unclear and more studies are needed to validate it.

The purpose of this study is to correlate the thickness of the shoulder axillary recess capsule measured by US with MRI signs of AC in patients with shoulder pain.

Methods

Eligibility criteria

This prospective study was approved by the Institutional Review Board of our hospital. The requirement for informed consent was waived. The authors prospectively selected 193 consecutive patients (141 women and 52 men; age range 20–89 years) referred to our hospital for shoulder US examination due to shoulder pain, from January 2015 to December 2016 (24 months). All subjects complained of pain with recent onset (0–9 months). Patients with rheumatic diseases were excluded (one patient).

Procedures and equipment

US examinations were performed by a board-certified radiologist with 20 years of experience in musculoskeletal ultrasonography. Studies were conducted using a Logic S8 system (General Electric Medical System, Milwaukee, Wisconsin, USA) with a linear array transducer ranging from 6 to 15 MHz and an iU22 (Philips Medical Solutions, Eindhoven, Netherlands) with a linear array transducer of 5–12 MHz.

Ultrasound: Measuring the joint capsule

All participants underwent routine shoulder US. Subsequently, while seated, patients' shoulders were placed in a position of external rotation and 90° abduction along with 90° forearm flexion–abduction external rotation (Figure 1(a)). In this position, the axillary recess becomes stretched, allowing measurement of the thickness of the capsular ligamentous/synovial complex and avoiding an overestimation resulted from neutral arm positioning. All patients were able to maintain this position. Transducer was positioned on the anterior axillary fold along with the diaphysis and humeral head (Figure 1(b)).

Figure 1.

(a) Ultrasound approach for the evaluation of the joint capsule thickness in the axillary recess with the abduction external rotation position. (b) Illustration displaying the positioning of the transducer for measurement of the joint capsule.

The joint capsule thickness value was considered a combination of capsular and synovial thicknesses of the axillary recess; it was characterized as a thin homogeneous hyperechoic band in contact with the humeral head cartilage (Figure 2(a) and (b)). Capsule thickness measurement was performed at the site where capsule was perpendicular to the ultrasound beam and adjacent to the humeral head in order to avoid anisotropy artifacts and to benefit from high contrast between the hyaline cartilage, joint capsule, and surrounding muscles. Calipers were positioned perpendicular to humeral head cartilage and the arithmetic average of three consecutive measurements was considered the final value (Figure 2(b)).

Figure 2.

(a) Detailed anatomy of the axillary recess with its capsular ligamentous/synovial complex (black arrow): synovia (blue line); inferior glenohumeral ligament (green line); joint capsule (orange line). Those components together form the axillary recess. pm – pectoralis major muscle; cb – coracobrachialis muscle; tm – teres major muscle. (b) Ultrasound in the abduction external rotation position showing the axillary recess complex with its capsular ligamentous/synovial complex as a single structure, represented by a hyperechoic line (between calipers) in contact with the articular cartilage of the humeral head (asterisk). pm – pectoralis major muscle; cb – coracobrachialis muscle; tm – teres major muscle.

After US, patients were separated into two groups: negative US (NUS) group, characterized by an axillary recess capsule thickness of 2.0 mm or less, and positive US (PUS), corresponding to patients with a capsule thickness greater than 2.0 mm. The 2.0 mm cutoff was estimated based on a previous study that measured joint capsule in asymptomatic patients.¹⁹

All patients from the PUS group underwent shoulder MRI (a total of 24 patients). In order to form a similar comparative group, 27 patients from the NUS group were randomly selected to undergo MRI scans.

MRI acquisition

The time between the US and MRI scans varied from 0 to 4 weeks. Exams were performed using 1.5 and 3.0 T MRI systems (Avanto, Siemens Medical Solutions, Erlangen, Germany; GE Optima 450W, GE Healthcare, Milwaukee, United States; Achieva, Philips Medical Systems, Best, Netherlands; Skyra, Siemens Medical Solutions, Erlangen, Germany; GE HDX, GE Healthcare, Milwaukee, United States) with a dedicated shoulder coil, without contrast injection and with section thickness of 3 mm.

MRI analysis was performed by a board-certified musculoskeletal radiologist with 10 years of experience, who was blinded to the clinical presentation and previous US examinations results. MRI findings evaluated included those previously reported in literature as accurate to the diagnosis of AC, as following: (1) Edema and thickening of the joint capsule in the axillary recess (axillary recess measurement was performed on coronal T2 fat-sat weighted sequence and considered abnormal when thickness was >4.0 mm).²⁰ (2) Signal abnormality on the rotator interval on an oblique sagittal T2-weighted image with fat saturation. (3) CHL thickness of > 4 mm (measured on the sagittal T1-weighted images). MRI was considered positive only in patients who presented at least two of the described criteria.

Authors correlated the axillary recess capsule thickness measured by US with MRI sings of AC.

Statistical analysis

Data are presented as absolute numbers (n) and relative (%) frequencies and were analyzed by Pearson's chi-square test. Odds ratios and their respective confidence intervals were calculated to determine the association of the variables with the outcome.

Continuous data were categorized by a receiver operating characteristic (ROC) curve through the point of greatest inflection of the curve.

Binary logistic regression was performed by adopting variables with p < 0.10 that were corrected for gender and age to adjust the odds value to the demographic variables. The Hosmer-Lemeshow test was performed to validate the logistic model.

Results

The total number of patients evaluated was 193, which included 141 women and 52 men between 20 and 89 years of age. Most of the subjects were between 40 and 69 years old. Demographic data of the PUS and NUS groups are summarized in Table 2. None of the patients in PUS group had diabetes mellitus, but this information was not collected from patients on NUS group. Two patients (2/24) in PUS group had a history of previous shoulder surgery.

Table 2.

Demographic data of the PUS and NUS groups.

	Total patients (n = 193)
	PUS group (n = 24)		NUS group (n = 169)		Pearson's chi-square test
	n	%	n	%	Sig.
Gender
F	12	50.0	129	76.3	0.007
M	12	50.0	40	23.7	0.007
Age
20–29	0	0.0	16	9.5	0.001
30–39	0	0.0	33	19.5
40–49	6	25.0	58	34.3
50–59	8	33.3	43	25.4
60–69	6	25.0	14	8.3
70–79	2	8.3	4	2.4
80–89	1	4.2	1	0.6
Ethnicity
Asian	5	20.9	8	4.8	0.030
Caucasian	18	75.0	140	82.9
African	1	4.2	21	12.4

Data in absolute numbers (n) and relative (%) frequencies. Pearson's chi-square test. PUS: positive ultrasound; NUS negative ultrasound.

Twenty-four patients (among the total of 193, representing 12%) had an axillary recess capsule thickness >2.0 mm (PUS group) (Figure 3). The axillary recess capsule of the 169 remaining patients (88%) had a thickness ≤ 2.0 mm (NUS group) and presented as a homogeneous hyperechoic band (Figure 2(b)).

Figure 3.

Adhesive capsulitis in a 56-year-old man. Axillary recess measurement (between calipers) shows marked thickening (3.4 mm), along with heterogeneous echogenicity.

In the PUS group, there was parity between men and women (50%). The mean age of the PUS group was 56 years, while the mean age in the NUS group was 45 years. Demographic and imaging data of the NUS group are summarized in Table 3.

Table 3.

Patients in the PUS group (capsular thickness > 2.0 mm)

Patient	Gender	Age (years)	Thickness^a (mm)	US findings	MRI
1	Male	49	3.8	−	+
2	Male	49	3.8	Supraspinatus Tendinopathy + bursitis	+
3	Male	43	3.8	Supraspinatus partial tear	+
4	Male	56	3.4	Supraspinatus partial tear	+
5	Male	40	3.5	PO rotator cuff Reconstruction	+
6	Male	82	2.3	Supraspinatus partial tear + bursitis	+
7	Male	55	2.6	PO supraspinatus Reconstruction + complete retear	+
8	Female	54	2.5	−	+
9	Female	41	5.1	−	+
10	Male	61	2.4	−	+
11	Female	61	3.1	−	+
12	Male	61	4.4	−	+
13	Female	64	2.8	Glenohumeral arthrosis	+
14	Female	72	3.0	PO supraspinatus	−
15	Female	46	2.2	−	+
16	Male	60	3.5	−	+
17	Female	70	5.6	Supraspinatus partial tear	+
18	Male	56	4.2	Biceps long head tendinopathy	+
19	Female	50	2.4	Supraspinatus tendinopathy	+
20	Female	50	2.9	Supraspinatus tendinopathy	+
21	Male	66	4.2	Supraspinatus tendinopathy	+
22	Male	54	2.8	Supraspinatus tendinopathy	+
23	Female	55	2.4	−	+
24	Female	41	5.0	−	+

US (−): negative ultrasonographic findings except capsular thickening; MR (−): negative MRI findings for AC; MR (+): MRI findings consistent with AC; PO: post-operative status.

Capsular thickness.

All 24 patients in the PUS group underwent a complementary shoulder MRI, and 23 of these had signs of AC (at least two of the three considered diagnostic criteria) (Figure 4). Only one patient in this group did not match the MRI criteria for AC.

Figure 4.

Shoulder MRI of the same patient from Figure 3 shows thickening and inflammatory changes in the axillary recess (a), in the rotator interval (b) and thickening of the coracohumeral ligament (c).

Twenty-seven patient from NUS group (27/169) were randomly assigned to perform shoulder MRI. None of them presented any of the diagnostic criteria for AC on MRI.

We observed that the performance of the ROC curve (Figure 5 and Table 4) for US discrimination between patients with MRI signs of AC in PUS group and NUS group had an accuracy value of 0.99, presenting an asymptotic value of p < 0.001. The cutoff value for the ROC curve with the most accurate thickness of the axillary recess for MRI signs of AC was 2.0 mm, corresponding to a sensitivity of 100% and a specificity of 96%.

Figure 5.

ROC curve for US with calculation of accuracy (area ± standard error) and asymptotic significance.

Table 4.

ROC curve for the USs with calculated accuracy (area ± standard error) and asymptotic significance.

Area	MSE	Asymptotic sig.	Confidence interval
Area	MSE	Asymptotic sig.	Inferior	Superior
0.99	0.014	<0.001	0.96	1.00

MSE: mean standard error.

Although there was a significant difference in age and gender between the PUS and NUS groups, with a higher mean age in the PUS group (Table 2), these factors did not influence the discriminatory ability of ultrasonography. A new ROC curve was generated to analyze the performance of the diagnostic test, with an average capsule thickness cutoff value of 2.0 mm in both age groups below and above 52 years old to exclude the age factor as a possible source of bias related to the thickening of the joint capsule. In addition, there was no significant difference between genders, as the performance of US for AC was similar for both males and females, according to Pearson's chi-square test (p < 0.001).

Logistic regression analysis was significant and proved the capacity of US for discriminating patients with and without AC, adjusted for gender and age. We observed an odds ratio of 4.16 for patients with a capsular thickness >2.0 mm, representing a strong diagnostic association as an independent variable.

To validate the diagnostic model, the Hosmer-Lemeshow test was used, with no significant difference between the predicted and observed values (p = 0.274).

Discussion

This study was able to demonstrate that a thickness greater than 2.0 mm in the axillary recess capsule measured on US correlated to MRI signs of AC. Additionally, we found that patients with a normal capsule thickness on US did not have signs of AC on MRI. These findings are in accordance with MRI studies that showed that capsular thickening and edema in the axillary recess have high accuracy for the diagnosis of AC as well as for determination of AC staging.^4,13,20

Only one patient in PUS group (with an axillary recess thickness of 3 mm) did not have signs of AC in MRI examination. We presume it was due to a recent surgical procedure to supraspinatus tendon reinsertion, which could cause a reactional capsule thickening (Table 3).

Kim et al.¹⁶ compared the US and MRI thickness of the axillary recess capsule in patients with clinical proved AC and showed that these measures correlates. They also demonstrated that the thickness of the axillary recess was significantly higher in the affected shoulder compared to the unaffected one. They found a threshold of 3.2 mm on US for the axillary recess in healthy patients (1.2 points higher than ours). This difference was probably due to the diverse arm positioning during the US examination, as they chose a shoulder abduction of 40° and elbow flexion of 90°. We chose a 90° shoulder abduction to allow some grade of stretch of the axillary recess and avoid overestimation of the measure by including soft tissue rather than the true capsule–synovial complex of the axillary recess.

Michelin et al.¹⁷ also evaluated the axillary recess by US and found same measurements as ours. In his study, asymptomatic patients had a capsule thickness lower than 2.0 mm with the shoulder positioned in maximum abduction (similar to our method), and all patients with AC presented an axillary recess thickness of 2.2 mm or higher.

Park et al.²¹ studied the axillary recess only in asymptomatic patients, in an attempt to create a basis for later studying the capsule thickness in AC shoulders. They found values ranging from 2.6 ± 0.5 mm to 3.1 ± 0.6 mm, using a different US approach and different measurement sites (measurements were conducted between the tip of the humeral head and the surgical neck).

Some other US studies have tried to establish the accuracy of US for diagnosing AC. Homsi et al.¹⁴ evaluated the accessibility and the CHL thickness in asymptomatic and AC patients, proving that a thickened CHL is suggestive of AC. However, such evaluation is many times limited since it may be difficult to obtain a reliable measurement of the CHL due to its complex anatomy without clearly definable margins.²²

Another author evaluated the rotator interval with power Doppler analysis¹⁵ and found that patients diagnosed with early-stage AC presented increased vascularity in the rotator interval. However, the rotator interval vascularity may also be altered by a biceps or subscapularis tendinopathy, presenting a similar appearance to that of fibro-inflammatory tissue related to AC.

Using the US axillary recess capsule assessment has the advantages of allowing an easy and feasible analysis and a bilateral comparative exploration. Also, arm abduction allows placement of the US transducer in the axilla and can tense the inferior glenohumeral ligament making the measurement of its thickness easier.

The most important clinical relevance of this study is the characterization of a valid correlation between the thickness of the axillary recess capsule measured on US and MRI signs considered highly suggestive of AC, allowing sonographers to consider this entity in patients with shoulder pain and a thickened axillary recess capsule. This data may avoid unnecessary MRI examinations which is many times demanded in patients with shoulder pain, especially when an atypical presentation. Besides, the maneuver suggested for the axillary recess measurements is of easy reproducibility for musculoskeletal sonographers, with easily recognizable anatomical references.

We recognize some limitations in this study. First, the lack of sonographer blinding, which could lead to a potential selection bias that may compromise the sensitivity and specificity of this study, once sonographers had access to patients complaints and clinical symptoms. This could be avoided by recording the key images to perform the measurements once they are anonymized, by a blinded sonographer.

Besides, a single radiologist performed US and MRI analysis and interobserver agreement of these findings was not attempted. Also, we did not correlate the US and MRI findings to clinical diagnosis of AC, once physical examination was not performed by an orthopedist. Our study is monocentric with a limited number of patients.

Although sonographic criteria for diagnosing AC have not been established yet, literature has offered many possibilities for its application. The axillary recess measurement may be a promising feature for such purpose. Also, we assume that a major study that combines the described signs of AC in US may show the most potential for concluding the real value of US in AC.

Conclusion

In patients with recent onset of shoulder pain, a thickness of greater than 2.0 mm of axillary recces capsule measured by US corresponds to MRI signs of AC with good sensitivity and specificity. Although promising, further studies are warranted to validate our results.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethics Approval

This prospective study was approved by the Institutional Review Board of Hospital Sírio-Libanês, Brazil, in January 2016.

Guarantor

RVL.

Contributors

RVL and RAS researched literature and conceived the study. GGC, ELB, and RSD designed the methodology. RAS and RVL conducted data collection. NSMRH conducted the data analysis. RVL wrote the first draft of the manuscript.

References

Harris

Bou-Haidar

Harris

. Adhesive capsulitis: review of imaging and treatment. J Med Imaging Radiat Oncol 2013; 57: 633–643.

Bunker

. Frozen shoulder: unravelling the enigma. Ann R Coll Surg Engl 1997; 79: 210–213.

Hand

Athanasou

Matthews

, et al. The pathology of frozen shoulder. J Bone Joint Surg Br 2007; 89: 928–932.

Ahn

Kang

, et al. Correlation between magnetic resonance imaging and clinical impairment in patients with adhesive capsulitis. Skeletal Radiol 2012; 41: 1301–1308.

Manske

Prohaska

. Diagnosis and management of adhesive capsulitis. Curr Rev Musculoskelet Med 2008; 1: 180–189.

Neviaser

. The frozen shoulder. Diagnosis and management. Clin Orthop Relat Res 1987; 223: 59–64. .

Neviaser

Hannafin

. Adhesive capsulitis: a review of current treatment. Am J Sports Med 2010; 38: 2346–2356.

Shaffer

Tibone

Kerlan

. Frozen shoulder. A long-term follow-up. J Bone Joint Surg Am 1992; 74: 738–746.

Connell

Padmanabhan

Buchbinder

. Adhesive capsulitis: role of MR imaging in differential diagnosis. Eur Radiol 2002; 12: 2100–2106.

10.

Chi

Kim

Long

, et al. Non-contrast MRI diagnosis of adhesive capsulitis of the shoulder. Clin Imaging 2017; 44: 46–50.

11.

Emig

Schweitzer

Karasick

, et al. Adhesive capsulitis of the shoulder: MR diagnosis. AJR Am J Roentgenol 1995; 164: 1457–1459.

12.

Gokalp

Algin

Yildirim

, et al. Adhesive capsulitis: contrast-enhanced shoulder MRI findings. J Med Imaging Radiat Oncol 2011; 55: 119–125.

13.

Gondim Teixeira

Balaj

Chanson

, et al. Adhesive capsulitis of the shoulder: value of inferior glenohumeral ligament signal changes on T2-weighted fat-saturated images. AJR Am J Roentgenol 2012; 198: W589–W596.

14.

Homsi

Bordalo-Rodrigues

da Silva

, et al. Ultrasound in adhesive capsulitis of the shoulder: is assessment of the coracohumeral ligament a valuable diagnostic tool?. Skeletal Radiol 2006; 35: 673–678.

15.

Walmsley

Osmotherly

Walker

, et al. Power Doppler ultrasonography in the early diagnosis of primary/idiopathic adhesive capsulitis: an exploratory study. J Manipulative Physiol Ther 2013; 36: 428–435.

16.

Kim

Cho

Sung

. Ultrasound measurements of axillary recess capsule thickness in unilateral frozen shoulder: study of correlation with MRI measurements. Skeletal Radiol 2018; 47: 1491–1497. .

17.

Michelin

Delarue

Duparc

, et al. Thickening of the inferior glenohumeral capsule: an ultrasound sign for shoulder capsular contracture. Eur Radiol 2013; 23: 2802–2806.

18.

Sofka

Ciavarra

Hannafin

, et al. Magnetic resonance imaging of adhesive capsulitis: correlation with clinical staging. HSS J 2008; 4: 164–169.

19.

Lefevre-Colau

Drapé

Fayad

, et al. Magnetic resonance imaging of shoulders with idiopathic adhesive capsulitis: reliability of measures. Eur Radiol 2005; 15: 2415–222.

20.

Park

Lee

Yoon

, et al. Evaluation of adhesive capsulitis of the shoulder with fat-suppressed T2-weighted MRI: association between clinical features and MRI findings. AJR Am J Roentgenol 2016; 207: 135–141.

21.

Park

Lee

Kwon

. Ultrasonographic measurement of the axillary recess thickness in an asymptomatic shoulder. Ultrasonography 2017; 36: 139–143.

22.

Edelson

Taitz

Grishkan

. The coracohumeral ligament. Anatomy of a substantial but neglected structure. J Bone Joint Surg Br 1991; 73: 150–153.