Ultrasound-Guided Glenohumeral Corticosteroid Injection for the Treatment of Adhesive Capsulitis of the Shoulder: The Role of Clinical Stage in Response to Treatment

Abstract

Background:

Treatment for idiopathic adhesive capsulitis of the shoulder remains controversial. Stages 1 to 2 reflect an inflammatory process supporting a rationale for intra-articular glenohumeral joint corticosteroid injection to treat synovial inflammation and prevent progression to capsular fibrosis.

Hypothesis:

We hypothesize that an intra-articular ultrasound-guided glenohumeral injection (USGI) of corticosteroid in patients diagnosed with Stage 1 or 2 idiopathic adhesive capsulitis will result in timely functional recovery and resolution of pain and stiffness.

Study Design:

Case series.

Level of Evidence:

Level 4.

Methods:

Patients with Stage 1 or 2 idiopathic adhesive capsulitis treated with an intra-articular corticosteroid injection were included. Patients were seen by a single physician and diagnosed using history and physical examination with careful attention to measurement of glenohumeral range of motion (ROM). A total of 61 patients met inclusion criteria. ROM measurements documented in the patient charts were recorded in forward flexion, abduction, internal rotation, and external rotation in neutral abduction. All ROM measurements were performed pre- and postinjection. All patients were treated with an USGI of local anesthetic and depomedrol. Recovery criteria were defined as forward flexion, abduction, and external rotation within 15° of the contralateral side and internal rotation within 3 spinous process levels of the contralateral side.

Results:

A total of 11 patients with Stage 1 and 50 patients with Stage 2 adhesive capsulitis comprised the final study cohort. Within the Stage 1 cohort, all 11 patients met recovery criteria for forward flexion and internal rotation (100%), 10 met recovery criteria for abduction (91%), and 8 met recovery criteria for external rotation (73%). Within the Stage 2 cohort, 31 patients met recovery criteria for forward flexion (62%), 30 met recovery criteria for abduction (60%), 36 met recovery criteria for internal rotation (72%), and 25 met recovery criteria for external rotation (50%). The difference in time to recovery in days was statistically significant in all ROM planes and was within 2 to 6 weeks for patients in Stage 1 and 7 to 10 weeks for Stage 2.

Conclusion:

USGI for early adhesive capsulitis allows patients to recover ROM more rapidly if performed early after onset of pain and stiffness.

Clinical Relevance:

These results stress the importance of recognition of idiopathic adhesive capsulitis in its early stages and subsequent intervention with an intra-articular glenohumeral corticosteroid injection.

Treatment of idiopathic adhesive capsulitis of the shoulder, colloquially referred to as frozen shoulder, remains controversial. There is a lack of consensus regarding optimal treatment as the results in the literature vary widely. The pathophysiology of adhesive capsulitis, described by Neviaser and Neviaser, ²² and refined by Hannafin et al, ⁹ is classified into 4 stages. Stage 1 involves synovial inflammation without significant capsular fibrosis; Stage 2 involves synovial inflammation, synovial scar, and early capsular fibrosis, while Stage 3 is characterized by capsular fibrosis without synovial inflammation. Based on the natural history of adhesive capsulitis, “benign neglect” has been recommended to allow the pathophysiology to progress through the stages and ultimately self-resolve.^2,20,24 Studies have demonstrated that long-term symptoms and disability may persist after this approach to treatment. ²⁹ Physical therapy, while utilized widely in clinical practice, has demonstrated mixed results.^2-6,13,18,32 Additional treatment interventions include nonsteroidal anti-inflammatory medications, ¹² oral corticosteroid, ² intra-articular corticosteroid injection,^{3,4,6,7,19,29,30,32} distension arthrography, ¹⁵ and closed manipulation.^1,17 Arthroscopic or open capsular release has demonstrated efficacy in shortening the time to recovery of adhesive capsulitis, but potential complications of surgery and anesthesia remain important considerations.^{16,23,25,28,31}

Many papers have been published examining the efficacy of treatment approaches for adhesive capsulitis of the shoulder; however, the majority fail to stratify the study cohorts into stage of disease at the time of treatment potentially confounding the results. The pathophysiology of adhesive capsulitis supports the rationale for intra-articular glenohumeral corticosteroid injection in patients presenting in Stages 1 and 2.^10,27 The aim of a corticosteroid injection is to reduce/ablate synovial inflammation and to prevent the development of capsular fibrosis, resulting in more timely improvement in range of motion (ROM) and decreased time to functional recovery. ¹⁰ We hypothesized that ultrasound-guided glenohumeral injection (USGI) of corticosteroid in patients diagnosed with Stage 1 or Stage 2 idiopathic adhesive capsulitis will result in a timely resolution of pain and stiffness and that the timing of injection with respect to stage will correlate with more rapid recovery of ROM.

Methods

All patients presenting to a single orthopaedic surgeon with Stage 1 or Stage 2 adhesive capsulitis treated with an USGI within a 6-year period (2013-2019) were identified within the electronic medical record. Patients were identified by International Statistical Classification of Diseases - ninth revision (ICD-9) code 726.0 and ICD-10 code M75.02. The diagnosis of idiopathic adhesive capsulitis was made based on history and clinical examination. Patients with Stages 1 and 2 disease presented with night pain, rest pain, and pain at end ROM. Patients with Stage 3 or 4 disease were distinguished clinically by the lack of rest or night pain and presentation as a stiff shoulder secondary to fibrosis. Physical examination findings and radiographs were evaluated to ensure that there was no concomitant rotator cuff tendinopathy, acromioclavicular joint pain, proximal biceps tendinosis, glenohumeral arthritis, or presence of neoplasm. Patients were excluded from this study if the etiology for adhesive capsulitis was postoperative or traumatic, or if they did not return for follow-up post-USGI. The senior author recommends USGI only for treatment of patients in the inflammatory stages of adhesive capsulitis; thus, patients with a clinical diagnosis of Stage 3 or 4 adhesive capsulitis were not treated with an USGI. The initial cohort contained 74 patients with Stage 1 or Stage 2 idiopathic adhesive capsulitis. There were 21 Stage 1 and 52 Stage 2 patients eligible for inclusion. Review of the medical record was performed under IRB No. 2019-1092. A total of 10 Stage 1 and 2 Stage 2 patients were eliminated secondary to lack of a follow-up examination after the USGI. A total of 61 patients met the inclusion criteria and were included in the final analysis. ROM measurements were recorded as documented as part of the standard patient evaluation in 4 planes: forward flexion, abduction, internal rotation (measured by having the patient place the thumb to the highest possible spinous process), and external rotation in neutral abduction. These ROM measurements were assessed with the patient standing and were measured by a single examiner without use of a goniometer. At the initial visit, supine isolated glenohumeral ROM was measured bilaterally, as subtle Stage 1 cases can be missed with the use of only standing ROM. ROM assessments were performed preinjection, at 2 weeks postinjection, and at all subsequent postinjection visits. All patients were offered treatment with an USGI containing 5 cc of 1% lidocaine, 3 cc of 0.25% Marcaine (bupivacaine hydrochloride), and 80 mg Depo-Medrol (methylprednisolone). Two patients (3%) required a second injection. Patients were given a home program for ROM in forward flexion, abduction and internal/external rotation to be initiated 1 week after injection or when rest/night pain had resolved. After the 2-week postinjection visit, patients continued the home-based ROM protocol if remaining ROM loss was minimal (Stage 1) or were referred for formal physical therapy if they still had significant loss of ROM (Stage 2). The stage of disease was confirmed by the initial response to treatment at the 2-week visit. Patients grouped as Stage 1 had resolution of night/rest pain with limited loss of motion. Patients grouped as Stage 2 had resolution of rest/night pain but had persistent loss of motion in all planes and pain at end ROM, reflecting stretch of the fibrotic capsule.

The exact time to recovery of motion was limited by variability in patient follow-up visits as improvement in motion may have been achieved more quickly in 1 plane compared with others. To address this variability, recovery criteria were defined a priori as forward flexion, abduction, and external rotation to within 15° of the contralateral side and internal rotation to within 3 spinal levels of the contralateral side.

Since time to recovery was the main outcome, cumulative survival functions were displayed graphically to show the proportion of patients who recovered over time (measured in days). The survival curves were compared between Stage 1 and Stage 2 patients using the log-rank test.

Results

A total of 11 patients with Stage 1 and 50 patients with Stage 2 adhesive capsulitis comprised the study cohort. The mean age was 55 years (SD, 8 years; range, 38-72 years) and 49 (80%) patients were female. Of the 61 patients, 11 (18%) had a thyroid disorder, 2 (3.3%) had diabetes mellitus (1 Type 1 and 1 Type 2), and 2 (3.3%) had Crohn’s disease. Further demographic data are described in Table 1.

Table 1.

Patient characteristics

Characteristic	Mean ± SD or N (%), (n = 61)
Age, y	55.2 ± 7.7 (range, 38-72)
Sex
Male	12 (20.0)
Female	49 (80.0)
Associated medical conditions
Thyroid disorder	11 (18.0)
Diabetes mellitus	2 (3.3)
Crohn’s disease	2 (3.3)
Stage 1	11
Days of symptoms	68.4 ± 22.6 (range, 30-99)
Stage 2	50
Days of symptoms	141.2 ± 68.7 (range, 25-324)

Within the Stage 1 cohort, all 11 patients met recovery criteria for forward flexion (100%), 10 patients met recovery criteria for abduction (91%), 11 patients met recovery criteria for internal rotation (100%), and 8 patients met recovery criteria for external rotation (73%). The mean duration of symptoms at initial presentation was 68.4 days (SD, 22.6 days; range, 30-99 days). The mean time to recovery in days for each ROM plane is listed in Table 2.

Table 2.

Mean days to recovery in each ROM plane

ROM Plane	No. of Patients (%)	Mean Days to Recovery (SD)	Range
Stage 1
Forward flexion	11 (100)	18.0 (4.1)	11-26
Abduction	10 (91)	18.0 (4.3)	11-26
External rotation	8 (73)	41.8 (46.3)	11-144
Internal rotation	11 (100)	23.7 (18.1)	11-77
Stage 2
Forward flexion	31 (62)	68.1 (67.9)	10-259
Abduction	30 (60)	66.5 (58.2)	12-259
External rotation	25 (50)	57.7 (42.9)	12-200
Internal rotation	36 (72)	54.3 (49.1)	10-207

ROM, range of motion.

In the Stage 2 cohort, 31 patients met recovery criteria for forward flexion (62%), 30 met recovery criteria for abduction (60%), 36 met recovery criteria for internal rotation (72%), and 25 met recovery criteria for external rotation (50%). The mean duration of symptoms at initial presentation was 141.2 days (SD, 68.7 days; range 25-324 days).

Time to recovery in days for each ROM plane (forward flexion, abduction, internal rotation, and external rotation) was compared between the Stage 1 and Stage 2 cohorts (Figures 1-4). The time to recovery was significantly increased in patients presenting with Stage 2 adhesive capsulitis reflecting the development of synovial and capsular fibrosis.

Figure 1.

Cumulative survival functions between Stage 1 and Stage 2 patients for forward flexion.

Figure 2.

Cumulative survival functions between Stage 1 and Stage 2 patients for abduction.

Figure 3.

Cumulative survival functions between Stage 1 and Stage 2 patients for internal rotation.

Figure 4.

Cumulative survival functions between Stage 1 and Stage 2 patients for external rotation.

The differences between the 2 groups in time to recovery of forward flexion (P < 0.01), abduction (P < 0.01), internal rotation (P < 0.01), and external rotation (P < 0.01) were statistically significant, with Stage 1 patients demonstrating a shorter time to recovery after injection. The mean time to recovery in days for each ROM plane for both Stage 1 and Stage 2 cohorts is depicted in Table 2.

Discussion

This study examined the impact of early intervention with USGI on the time to recovery of ROM in patients with Stage 1 and Stage 2 adhesive capsulitis. The results of this study support our initial hypothesis that an USGI of corticosteroid in patients diagnosed with Stage 1 or Stage 2 idiopathic adhesive capsulitis results in timely functional recovery and resolution of pain and stiffness. Patients with Stage 1 disease recovered forward flexion ROM on average 18.0 days after injection (SD, 4.1 days; range, 11-26 days), whereas patients with Stage 2 disease recovered forward flexion on average 68.1 days after injection (SD, 67.9 days; range, 10-259 days). Stage 1 patients reported symptoms for 68.4 days on average, while those with Stage 2 disease reported an average of 141.2 days of symptoms.

Adhesive capsulitis of the shoulder is a condition of varying severity characterized by a painful and gradual loss of both active and passive glenohumeral motion with limited radiographic findings other than signs of osteopenia. The staging system for adhesive capsulitis was first described by Neviaser and Neviaser ²² in 1987 and refined by Hannafin et al ⁹ by correlating clinical, arthroscopic, and histological findings to better delineate the Stages 1 to 3. In Stage 1, patients present with rest and night pain and can demonstrate subtle or significant loss of ROM. An erythematous synovial inflammatory reaction without adhesions or capsular contracture is seen arthroscopically, and histology demonstrates an inflammatory synovitis with normal capsular tissue. In Stage 2, patients also present with rest and night pain and demonstrate more significant limitation of ROM with restrictions in forward flexion, abduction, internal, and external rotation. Arthroscopic assessment demonstrates extensive synovitis, with global loss of capsular volume. Histology demonstrates synovial inflammation and perivascular scarring with concomitant capsular fibrosis. The third stage is distinguished clinically by profound stiffness, pain associated with end ROM and the lack of rest and night pain. Significant loss of motion is a hallmark of Stage 3, and there is no improvement in ROM under anesthesia. Arthroscopic evaluation demonstrates no significant inflammation, and histology demonstrates hypercellular, collagenous tissue with a thin synovial layer and features common to other fibrosing conditions. An USGI performed at the early stages of this disease process provides chemical ablation of the synovitis which is a characteristic pathologic finding in Stage 1 and Stage 2,^8,18,19 thus limiting the progressive development of capsular fibrosis and shortening the natural history of adhesive capsulitis.

Understanding the pathophysiology of adhesive capsulitis, which is characterized by stage dependent inflammation and fibrosis, is necessary to devise an adequate treatment plan. The current literature varies extensively when describing the response to suggested treatments, which is likely attributed to a lack of a stage-based approach that accounts for the pathophysiology of each stage. To our knowledge, the current study is the first study to separate cohorts based on stage and to compare time to recovery between the 2 groups after ultrasound-guided intra-articular corticosteroid injection.

In this study, all the patients treated with USGI in Stage 1 recovered their forward flexion and internal rotation, with all but 1 patient (91%) fully recovering their abduction and 73% recovering their external rotation. Excluding external rotation, they met the criteria for successful treatment within 3 weeks of their injection (mean, 18.1-23.7 days). External rotation was recovered on an average of 41.8 days. Patients with Stage 2 adhesive capsulitis had a slower time to recovery of over 7 weeks (mean, 54.3-68.1 days) and a significantly lower percentage of patients met recovery criteria within each plane of ROM (forward flexion 62%, abduction 60%, external rotation 50%, internal rotation 72%).

Because the natural course of adhesive capsulitis is often self-limiting in nature, “benign neglect” has been a recommended conservative treatment approach. However, long-term follow-up patients who were treated with benign neglect demonstrated that 50% of patients reported persistent shoulder pain, stiffness, or a combination of both, and 60% demonstrated long-lasting limitation in ROM. ²⁹ The most common treatment for adhesive capsulitis is physical therapy. A prospective study of 75 patients with Stage 2 adhesive capsulitis treated with specific 4-direction shoulder-stretching exercise program concluded that most patients with Stage 2 adhesive capsulitis achieved a satisfactory outcome. ⁸ O’Kane et al ²⁴ demonstrated that a simple physical therapy home program for frozen shoulder improved patients’ assessment of shoulder function and health status based on questionnaires for simple shoulder test and short-form health survey (subjective). On the contrary, several studies have demonstrated no difference between patients treated with physical therapy and no treatment.^3,4 Again, these studies fail to state explicitly the stage of disease at the time of treatment, making it difficult to accurately assess the impact of physical therapy as a treatment modality. In addition, “no-treatment” controls are not consistently “no-treatment” as patients are engaging in home stretching programs and/or pendulum exercises. ⁵ Arthroscopic debridement and capsular release for adhesive capsulitis has demonstrated symptom improvement; however, the risks and complications associated with any surgical procedure make this a less desirable first-line treatment option.^{5,16,23,25,28,31}

The published literature supports the use of an intra-articular corticosteroid injection in the treatment of adhesive capsulitis. van der Windt et al ³² compared corticosteroid injection to physical therapy and found that 77% of patients who underwent injection had successful resolution of adhesive capsulitis compared with only 46% of patients who received physical therapy alone. Bulgen et al ³ compared 4 treatment groups: (1) intra-articular injection, (2) mobilization with physiotherapist, (3) ice treatment after proprioceptive exercises, and (4) no treatment. All groups reported improved pain at 6 months, with the intra-articular steroid injection treatment group demonstrating the most significant improvement in ROM at 4 weeks posttreatment. Again, the stage of disease was not specified. ³ A placebo-controlled trial demonstrated that a single, fluoroscopically guided injection of 40 mg triamcinolone hexactonide produced significantly lower Shoulder Pain and Disability Index scores compared with placebo injection and physical therapy, or placebo injection alone. However, patients were divided into acute and chronic adhesive capsulitis and not classified by stage. ⁴ In a retrospective study of 130 patients, the efficacy of intra-articular injections of hydrocortisone was correlated inversely with duration of symptoms. ¹⁴ A retrospective case series of Stage 1 and Stage 2 adhesive capsulitis patients treated with a nonultrasound-guided, office-based depomedrol injection demonstrated time to recovery of 6 weeks and 7 months, respectively. ¹⁹ The accuracy of “blind” versus ultrasound-guided intra-articular injections of the shoulder was examined comparing anterior and posterior approaches. The anterior approach was found to be 94% accurate, while the posterior approach was 78% accurate. ²⁶ The consensus regarding accuracy of non-USGI remains unclear when compared with USGI. For this reason, an ultrasound-guided intra-articular glenohumeral injection was used by the senior author.

Further studies assessing treatment modalities for adhesive capsulitis of the shoulder with attention to disease stage are necessary to reach consensus on standard of care treatment. A meta-analysis of 9 randomized controlled trials with a total of 453 patients demonstrated steroid injections to be equivalent in effectiveness to physical therapy. ³³ Multiple randomized controlled trials have demonstrated that when steroid injections are used adjunctively with physical therapy, a greater improvement in symptoms is observed when compared with physical therapy alone.^3,4,14 The current study demonstrates a significant shortening of the duration of symptoms after corticosteroid injection when grouped by disease stage. It is essential that future studies define patients by presumptive stage and duration of symptoms at presentation. It is likely that as the pathophysiology of adhesive capsulitis changes across stages, different treatment modalities may exhibit variable efficacy.

This study has several limitations. This was a retrospective analysis of prospectively collected data, which inherently introduces concerns for reliability of chart review. There was no control group as all patients in Stage 1 or Stage 2 were offered an USGI for treatment. The presence of comorbidities, such as endocrine disorders, were not assessed as independent variables, secondary to lack of statistical power. Demographically, 18% of the patients in this study had concomitant thyroid disease and 3.2% were diabetic, both known risk factors for the diagnosis of adhesive capsulitis.^11,33 The lack of follow-up on 10 Stage 1 and 2 Stage 2 patients limited the final cohort. Interestingly, patients treated in Stage 1 often have full resolution of symptoms within 2 weeks after injection, which may explain the lack of follow-up. In addition, the study cohort was dominated by Stage 2 patients. As discussed previously, it is difficult to define recovery for this condition. Recovery of ROM was utilized as the outcome in each of the 4 planes of motion and converted to a binary variable (ie, recovery or no recovery). The recovery criteria were defined a priori based on clinical relevance. Finally, it was difficult to evaluate ROM at consistent intervals as patients’ follow-up appointments varied in time intervals and it is likely that some patients may have recovered sooner than the time they presented for their follow-up evaluation.

This retrospective, single surgeon, cohort study of 61 patients with Stage 1 and 2 adhesive capsulitis managed with USGI demonstrates that early diagnosis and subsequent intervention lends itself to a faster recovery of ROM. Patients with Stage 1 disease managed with an USGI met recovery criteria in less than half the time it took their Stage 2 counterparts to meet the same criteria. In addition, the percentage of patients who met the ROM recovery criteria after USGI was statistically significantly higher in the patients with Stage 1 adhesive capsulitis versus those with Stage 2. The results of this study emphasize the importance of early detection and diagnosis of adhesive capsulitis as presentation at Stage 1 can be subtle. Evaluation of the glenohumeral ROM with the patient supine provides scapular stabilization and can be helpful in the diagnosis of early adhesive capsulitis. The results of this study support and advocate for the early use of USGI for patients with adhesive capsulitis to facilitate rapid recovery of ROM and return to normal function by shortening the natural history of adhesive capsulitis. It will be important in future studies to assess patient-reported outcome measures in addition to the cost-benefit of prompt treatment for adhesive capsulitis.