A stratified investigation of a single injection of extended-release corticosteroid for symptomatic relief in patients with idiopathic adhesive capsulitis of the shoulder

INTRODUCTION

The clinical examination of idiopathic adhesive capsulitis is a progressive restriction of passive glenohumeral motion, driven by a cascade of capsular inflammation and subsequent fibrotic contracture.^[1] It occurs in approximately 2–5% of the population. While the true etiology of the condition remains unknown, it has been proposed that synovitis triggers a fibrotic cascade.^[2] Systemic metabolic factors, most notably diabetes mellitus and thyroid disorders, are associated with an increased prevalence of frozen shoulder.^[1,2]

Non-operative protocols prioritize physical therapy and home-based stretching.^[1] Oral corticosteroids, intra-articular corticosteroid injections, hydrodilatation, and non-steroidal anti-inflammatory medications are often used as adjuncts to therapy. Arthroscopy with lysis of adhesions and/or manipulation under anesthesia may be needed for patients who fail standard steroid injections.^[1,2]

Intra-articular injection of corticosteroids for frozen shoulder is superior to treatment with oral cortisone^[3,4] and at least equivalent to isolated manipulation under anesthesia.^[5] Injections have been shown to decrease pain and result in earlier return of passive range of motion (ROM).^[6-12] Lower doses of steroid injections have demonstrated similar efficacy as a higher dose treatment.^[9,12] Image-guided injections have also demonstrated better outcomes compared to landmark-guided injections for shoulder pain.^[13]

If an injection does not provide adequate duration of relief, potential next-step treatment options include repeat intraarticular injection or surgical intervention. Either option incurs further healthcare costs and potential morbidity to the patient. It could be of great value to achieve prolonged efficacy with extended-release intra-articular steroid injection for adhesive capsulitis. Previous studies on extended-release corticosteroid injections have shown promise for treating knee osteoarthritis (OA).^[14-18] Long-acting steroids, which function by slow release from a biodegradable microsphere, may produce a longer duration of efficacy and avoid the need for costly repeat injections or surgical intervention. An additional potential benefit is less elevation of peripheral blood glucose in diabetics, a known complication of traditional steroid injections that has been demonstrated to be mitigated using a sustained-release formulation.^[19]

This study examined the effect of an extended-release corticosteroid injection in the treatment of idiopathic adhesive capsulitis of the shoulder. The primary objective of this study was to assess patient-reported pain scores through 1-year follow-up following a single, image-guided injection in patients with idiopathic adhesive capsulitis of the shoulder. Secondary goals of the study were to assess patient-reported outcome (PRO) scores and passive ROM through 1-year follow-up. Subgroup analysis was performed for patients with and without diabetes mellitus. Based on prior literature regarding steroid injections for this condition, we hypothesized that there would be a significant reduction in pain at all-time points following the injection.

MATERIALS AND METHODS

This was a phase 2, stratified, prospective, non-blinded single-arm study of patients with idiopathic adhesive capsulitis who received a single, image-guided intra-articular extended-release corticosteroid injection. The study protocol was approved by the Institutional Review Board. All patients provided written informed consent for the prospective study.

RESULTS

A total of 39 patients were enrolled in the study. The cohort had a mean age of 55 ± 9 years and a mean body mass index of 29 ± 6 kg/m². Only one patient had prior ipsilateral shoulder surgery: an arthroscopic debridement performed more than 20 years before the study period. Most patients were female (72%) and white (85%), with 38% of patients diagnosed with diabetes [Table 1]. Thirty-two (82%) patients in the study cohort achieved adequate symptom resolution after extended-release corticosteroid injection. Five (13%) patients underwent subsequent injections for recurrent adhesive capsulitis symptoms during the study period, and two (5%) patients underwent surgery on the study shoulder. One patient underwent an arthroscopic lysis of adhesions with manipulation under anesthesia 5 months after her Zilretta^® injection, and the second patient underwent arthroscopic rotator cuff repair 8 months after his Zilretta^® injection.

Visual analog pain scale

The VAS pain score improved from a mean of 5 at the time of injection to 1.4 by 6-week post-injection (p < 0.01). This significant improvement in VAS pain score persisted at 3, 6, and 12-month follow-ups (p > 0.05 for all comparisons) relative to baseline, indicating sustained improvement [Figure 1].

Close

Figure 1: Visual analog pain scale pain scores demonstrated significant improvement from baseline by 6 weeks post-injection that was sustained out to 12 months (1.5 months: n = 35, SD = 1.7; 3 months: n = 29, SD = 1.5; 6 months: n = 27, SD = 2.4; 12 months: n = 18, SD = 1.1). SD: Standard deviation; VAS: Visual analog pain scale pain scores.

Export to PPT

American Shoulder and Elbow score

The ASES score improved from a mean of 43 at the time of injection to a mean of 80 post-injection (p < 0.01). There were no significant changes in the mean ASES score after the 6-week follow-up (p > 0.05 for all comparisons), indicating continued improvement through 12-month follow-up [Figure 2].

Close

Figure 2: American shoulder and elbow score scores demonstrated significant improvement from baseline by 6-week post-injection that was sustained out to 12 months (1.5 months: n = 35, SD = 17; 3 months: n = 29, SD = 14; 6 months: n = 27, SD = 20; 12 months: n = 18, SD = 15). SD: Standard deviation; ASES: American shoulder and elbow score scores.

Export to PPT

Range of motion

Forward elevation improved from a mean of 117° at the time of injection to 173° at 3-month post-injection (p < 0.01). There were no significant changes in mean forward elevation at the 3-month follow-up (p > 0.05 for all comparisons), indicating sustained improvement. Abduction improved from a mean of 98° at the time of injection to 162° at 3-month post-injection (p < 0.01). There were no significant changes in mean abduction at the 3-month follow-up (p > 0.05 for all comparisons). External rotation with the arm adducted to the side improved from a mean of 32° at the time of injection to 59° at 6-week post-injection (p < 0.01). There were no significant changes in the mean external rotation with the arm adducted after the 6-week follow-up (p > 0.05, all comparisons), indicating a sustained improvement [Figure 3]. External rotation in 90° of abduction, as well as internal rotation, did not demonstrate statistically significant changes.

Close

Figure 3: Measurements of passive range of motion demonstrated sustained significant improvements in external rotation with the arm adducted by 6-week post-injection, in forward flexion by 3-month post-injection, and abduction by 3-month post-injection. FF: Forward flexion; ABD: Abduction; ER0: External rotation with the arm adducted to 0°; ER90: External rotation with the arm abducted to 90°; ROM: Range of motion.

Export to PPT

DISCUSSION

In this study, 82% of patients who underwent a single extended-release corticosteroid injection for acute, idiopathic adhesive capsulitis of the shoulder did not require any further treatment for their symptoms. At 6 weeks after a single extended-release corticosteroid injection, there was pain relief that was maintained through a 1-year follow-up. PROMs improved significantly by 6 weeks post-injection and remained sustained at 1-year follow-up. Passive external rotation with the arm adducted improved from 32° at the time of injection to 59° at 6-week post-injection. Passive forward elevation and abduction were improved by 3 months after injection (117– 173° and 98–162°, respectively). Outcomes were comparable in both diabetic and non-diabetic patient subgroups.

Mean VAS pain scores at baseline in this study cohort were consistent with those reported in studies investigating intraarticular steroid use for adhesive capsulitis.^[4,9,20] The current study revealed a rapid reduction in VAS scores within the first 6 weeks and the preservation of these improvements through the 1-year mark. While direct comparisons are difficult, the early follow-up results appear at least non-inferior to those reported in the existing literature on standard intra-articular corticosteroid injections. Yoon et al. reported a VAS pain improvement from 5.2 to 2.4 at 12-week follow-up in a cohort of 20 patients who received a 40 mg triamcinolone acetonide injection for adhesive capsulitis.^[9] In a prospective, randomized clinical trial, Ranalletta et al. reported a reduction in VAS scores from 7.6 at baseline to 1.1 at 12-week follow-up in a cohort of 37 patients receiving betamethasone.^[4] However, both of these studies are limited by their short-term follow-up. In a few studies that have examined longer-term follow-up (beyond 3 months) after intra-articular corticosteroid use, efficacy wanes over time.^[21-24] Ryans et al. showed that a standard intra-articular corticosteroid injection was associated with a significant improvement in VAS scores compared to physical therapy alone at 6-week follow-up, but by 16-week follow-up, there was no significant difference between groups.^[24] Bal et al. and Van der Windt et al. also reported similar findings, with early improvements in pain after injection but minimal differences between intra-articular steroid and physical therapy alone at 12 and 26 weeks, respectively.^[22,23] The current study cohort had both quick pain relief and sustained relief at longer-term follow-up, which suggests that further studies investigating possible non-inferiority (or superiority) of extended-release corticosteroid injection are warranted.

PROMs following injection were analyzed and improved significantly from 43 to 80 by 6-week post-injection, which is similar to the effects of standard intra-articular steroid. For example, one clinical trial reported an improvement in ASES scores from a mean baseline of 42–81 at 4 weeks of follow-up after betamethasone injection.^[4] That study, however, reported only outcome scores at a final follow-up of 12 weeks. The findings of Sun et al. suggest that standard steroid injection benefits wane and are equivalent to therapy by 24 weeks. Using an extended-release formulation, our cohort maintained improvements through the 1-year follow-up.^[10] The results of the current study showed that the early benefit of the extended-release corticosteroid injection persisted, perhaps related to the extended-release nature of this novel medication compared with standard corticosteroids.

Passive ROM was analyzed in this study with significant improvements in passive external rotation with the arm adducted at 6 weeks, passive forward elevation at 3 months, and passive abduction at 3 months. The results in the present study appear consistent with other reported data. For instance, Ryans et al. reported an approximately 20° improvement in passive external rotation at 16-week post-injection.^[24] The current study demonstrated an approximately 30° improvement in passive external rotation during the 3–4-month follow-up period.

Patients with a prior history of diabetes mellitus were specifically analyzed, and this cohort demonstrated significant improvements in mean VAS pain scores, mean ASES scores, mean forward flexion, and mean external rotation with the arm adducted at 6 weeks post-injection. Mean abduction was also significantly improved in this cohort by 3-month post-injection. All of these improvements were sustained for 1 year. Given that adhesive capsulitis affects up to 36% of diabetic patients (far exceeding the 2–5% reported in the general population), evaluating treatments specifically in this subgroup is of great clinical value.^[25,26] Adhesive capsulitis also tends to be more severe and recalcitrant to conservative management in the diabetic population. Erickson et al. showed that diabetic patients presented with worse forward elevation and failed conservative management more often.^[27] Those authors further demonstrated that adhesive capsulitis was less likely to resolve after one standard corticosteroid injection and required capsular release more often in the diabetic cohort.^[27] It is therefore of paramount importance to report on the outcomes of extended-release corticosteroid injection in this patient population. Perhaps unsurprisingly, non-diabetics showed greater improvement in pain and ASES scores at certain time points. Another outcome that should be studied in the future with glenohumeral joint steroid injections is post-injection fluctuations in blood sugar. There is the potential that extended-release medications could limit the bolus effect of standard steroid injections, but this will require additional investigation.

The non-blinded nature of this study introduced bias into PROMs, such as pain and ASES scores. Passive ROM recordings were unaffected by this, since they are unbiased, objective data points for patient improvement. The relatively small patient cohort made it difficult to perform robust subgroup analyses among diabetic patients due to insufficient statistical power. There were four patients who received shoulder steroid injections before the study period. These injections were performed at outside institutions, and thus, the rationale for these prior injections was limited. The cohort size also limited the ability to detect rare adverse reactions to the injection, although it is Food and Drug Administration-approved for the treatment of knee OA. During the follow-up period, there was a loss of follow-up for unknown reasons that could have impacted the results. This makes the current study susceptible to attrition bias toward either positive or negative outcomes. Patients may not have returned for evaluation for many possible reasons, such as an asymptomatic shoulder or worsening symptoms that prompt a second opinion. Finally, the current study lacked a control group, which makes it impossible to determine whether extended-release corticosteroid injection is superior to standard intra-articular steroid injection. In addition, without a control group, it is difficult to determine how extended-release may have affected the natural history of frozen shoulder, as it may improve over time without intervention.^[28] However, this was not the goal of the phase-2 study, which focused on efficacy and safety.

CONCLUSION

A single extended-release corticosteroid injection was associated with sustained pain relief, improvement in PROs, and increased ROM in patients with idiopathic adhesive capsulitis of the shoulder.