Pathophysiology of adhesive capsulitis of shoulder and the physiological effects of hyaluronan

Abstract

Adhesive capsulitis, commonly described as “frozen shoulder,” is experienced as chronic soreness and discomfort accompanied by an increasing inability to use and manipulate the joint, which can cause disability and interfere with normal activities. The onset of the condition is sudden and the origins obscure. The more we know about how this condition arises and evolves in the human body, the better the medical profession can provide effective management and amelioration. In examining the pathology and pathophysiology of adhesive capsulitis, we found recent evidence that intra-articular use of hyaluronan can provide effective relief. In this letter, we summarized this research to obtain a better understanding of the pathophysiology of adhesive capsulitis and the effects of hyaluronan.

Introduction

Adhesive capsulitis (AC) occurs in 3%–5% of the general population every year. The onset of the condition has been attributed to diabetes mellitus, thyroid dysfunction, Dupuytren’s contractures, autoimmune disease, and the treatment of breast cancer.

The general treatment and management of the condition involves the use of non-steroidal anti-inflammatories, corticosteroid injections into the affected area, physiotherapy, hydrodilation, arthroscopic release, and manipulation while anesthetized. ¹ Recent studies have also suggested that hyaluronan (HA) injected into the site can help recovery. ² Our objective, therefore, is to study the pathology and pathophysiology of AC and the clinical and physiological effects of applying HA.

AC

Pathology of AC

Codman first used the term “frozen shoulder,” describing it as a condition that was equally hard to categorize, to treat, and to illustrate. After about 80 years, that is still largely the case. Eleven years after Codman, Neviaser used the more technical term “adhesive capsulitis,” highlighting the inflammation and changes to the capsule or bursa between the scapula and the humerus, and described its pathological progress as a four-stage process. ¹

In the first stage, the patient feels a progressive attack of pain in the deltoid muscle area, which is usually more acute at night. There may be some restriction of movement, but this disappears when an anesthetic is injected into the joint. Within the joint there are signs of synovitis without adhesions or contractures. Biopsy of the joint capsule reveals the presence of unusual inflammatory cells, hypervascular, hypertrophic inflammation of the synovial membrane, and normal capsular tissue.

In the second stage, the sufferer is aware of stiffness in the joint. Arthroscopy shows thickening of the synovial membrane and adherence to other soft tissues, and the articulative function of the shoulder joint is less effective. The pain becomes chronic and, again, is worse at night. The patient now finds it difficult to move the joint forward or away from the body or to rotate it. The tissues are subject to hypertrophic, hypervascular synovitis with perivascular and subsynovial scar formation.

In the third stage, also called the maturation stage, there is extensive loss of movement in the joint and any movement is accompanied by pain. The stiffness is the most difficult sensation to bear. By now, the swollen and adhering connecting tissues do not allow the joint to function, as evidenced by capsular biopsy of the dense hypercellular collagenous tissue, especially at the front of the capsule.

When the condition enters the fourth, or chronic, stage, the inelasticity remains but is accompanied by little or no pain as the inflammation has run its course. Arthroscopy shows that the scar tissue and adhesions are now fully developed and conceal the structure within the joint.

While it is convenient to describe this as a staged process, in reality it is likely to present as a flow, and there is widespread disagreement among professionals as to the outcome of the condition and its long-term effects.

Pathophysiology of AC

There is widespread agreement that AC involves both inflammation and thickening and scarring of the connective tissue. Cyclooxygenases (COXs) contribute to inflammation and the breakdown of the structure of peripheral tissues. Using reverse transcription polymerase chain reaction (RT-PCR) and immunohistochemistry (IHC), Lho et al. ³ showed that AC sufferers had higher levels of inflammatory cytokines, including interleukin (IL)-1α, IL-1β, tumor necrosis factor (TNF)-α, and COX-1 and COX-2, in capsular and subacromial bursal tissue than those of the controls. Analysis of glenohumeral capsular tissue also shows that patients with AC have raised levels of intercellular adhesion molecule-1 (ICAM-1), a transmembrane protein on endothelial cells and leukocytes that assists leukocyte endothelial transmigration in capsular tissue, synovial fluid, and serum. Patients with diabetes mellitus also have raised ICAM-1 levels. ⁴ Neural changes that occur in AC indicate raised levels of a number of immune-reactive neuronal proteins, including growth-associated protein 43, protein gene product 9.5, and nerve growth factor receptor (P75), in the anterosuperior joint capsule. ⁵ Hence, we can conclude that neoangiogenesis takes place in AC together with neoinnervation—the latter probably accounting for the extreme pain associated with the condition.

The microscopic examination of tissue from AC sufferers reveals fibroblasts mixed with type I and type III collagen, which has led to AC generally being regarded as a fibrotic disorder like Dupuytren’s disease. The fibroblasts change into smooth muscle phenotype (myofibroblasts), which appears to cause the capsular contraction. The advancement of the condition seems to be due to the presence of adhesion-related cytokines, that is, transforming growth factor β (TGF-β) and platelet-derived growth factor (PDGF). TGF-β is associated with fibrosis and accumulation of a dense matrix of type I and type III collagen within the capsule and with increased PDGF, ⁶ and some research suggests that it is the stimulant for AC. Another study by Watson et al. indicated that excessive TGF-β1, injected into the joints of rats using an adenovirus vector, led to the development of AC after 5–10 days. Cell cultures were carried out on both contracted and normal patient tissues followed by blocking of the actions of TGF-β and PDGF, both of which were indicated at raised levels in the contracted tissue. Blocking led to a decrease in the formation, spread, and chemotaxis of collagen. In another study, Bunker et al. compared the occurrence of tissue sample matrix metalloproteinases (MMPs), which play a role in scar tissue remodeling, and their inhibitors in AC sufferers and a control group. The absence of MMP-14 was noted in the AC group. The remodeling of the matrix is controlled by MMPs. ⁷ Because MMP-14 is an activator of MMP-2, which is involved in collagen degradation, the absence of the former may result in excess production of collagen rather than its disintegration. Another study examined the serum levels of proteins related to AC, MMP-1, MMP-2, tissue inhibitor of metalloproteinase (TIMP)-1, TIMP-2, and TGF-β1. MMP-1 and MMP-2 levels were significantly lower, and TIMP-1, TIMP-2, and TGF-β1 levels were significantly higher in the AC group. ⁸ From this evidence, we can hypothesize that AC is caused by a breakdown in the interplay between the degradation, remodeling, and regeneration of extracellular matrix (ECM) tissue. Raised levels of extracellular sinal-regulated kinases (ERK), the Jun N-terminal kinases (JNK), nuclear factor kappa B (NF-κB), CD29, and vascular endothelial growth factor have also been found in local tissues obtained from AC sufferers. ⁹

Recent studies have attempted to establish whether the breakdown of molecular structures is associated with known risk factors and genetic liability to contract AC. A study by Xu et al. ¹⁰ led to the conclusion that single nucleotide polymorphisms in the MMP-3 rs650108 variant were significantly associated with increased risk of AC.

From the above studies, we can deduce that AC is a pathophysiological process involving the secretion of multiple cytokines such as IL, COX, and TGF, leading to inflammation and fibrotic dysfunction (Table 1).

Table 1.

The pathophysiological research of adhesive capsulitis.

Research	Number of cases	Sample	Histological and molecular result
Lho et al. 3	14	Joint capsules and subacromial bursae	IL-1α↑, IL-1β↑, (TNF)-α↑, COX-1↑, and COX-2↑
Kim et al. 4	17	Glenohumeral capsular tissue and synovial fluid	ICAM-1↑
Xu et al. 5	8	Shoulder capsular samples	P75↑, PGP9.5↑, and GAP43↑
Rodeo et al. 6	19	Shoulder capsule and synovium	TGF-β↑, PDGF↑, and HGF↑
Bunker et al. 7	14	Tissues from rotator interval	IL-1α↑, IL-6↑, PDGF-α↑, PDGF-β↑, TNF-β↑, basic-FGF↑, MMP-1↓, MMP-2↓, TIMP-1↑
Lubis and Lubis 8	50	Serum	MMP-1↓and MMP-2↓, while TIMP-1↑, TIMP-2↑, TGF-β1↑
Kanbe et al. 9	10	Synovial tissues between the long head of the biceps tendon and the rotator interval	IL-6↑, MMP-3↑, and VEGF↑
Xu et al. 10	112	Peripheral blood	MMP-3 rs650108↑

IL: interleukin; TNF: tumor necrosis factor; COX: cyclooxygenase; ICAM: intercellular adhesion molecule; P75: nerve growth factor receptor P75; PGP9.5: nerves positive for nerve growth factor receptor 9.5; GAP43: growth-associated protein 43; TGF-β: transforming growth factor-beta; PDGF: platelet-derived growth factor; HGF: hepatocyte growth factor; FGF: fibroblast growth factors; MMP: matrix metalloproteinase; TIMP: tissue inhibitor of metalloproteinase; VEGF: vascular endothelial growth factor; ↑: increased; ↓: decreased.

HA

Clinical research into HA

Recent research suggests that HA treatment can reduce pain and improve the mobility of the shoulder joint. In a clinical experiment, AC patients received either injections of HA or intra-articular steroids, or physical therapy. The severity of pain, degree of mobility, and functional considerations were measured before treatment and at 15 days and 3 months afterward. ² By the third month, the condition of all of the treated groups had improved significantly compared with the no-treatment control group. ² In another study, Jang et al. compared the effects of intra-articular HA and steroid injection in patients with hemiplegic shoulder pain after stroke. Hemiplegic shoulder pain is classified as AC. Shoulder pain was eased and mobility significantly improved in both groups. The steroid injections were more effective in pain control in the early stages of treatment, but there was no significant difference at 4 and 8 weeks. The HA group reported significantly less pain at night. ¹¹

Sodium HA (20 mg) administered by injection together with triamcinolone acetonide (20 mg) in addition to physiotherapy appeared to provide a greater degree of pain relief and mobility in AC than injected triamcinolone in association with physiotherapy. A clinical trial by Park et al. compared intra-articular administration of HA at the same time as capsular distention (achieved by 0.5% lidocaine, 18 mL) with intra-articular corticosteroid injection. This trial claimed good outcomes in terms of pain (Verbal Numeric Scale), function/disability (Shoulder Pain and Disability Index), and range of motion (passive). Passive external rotation was significantly improved in the HA (combined with capsular distention) group at 6 weeks after injection. ¹² HA may be more effective in combination with capsular distention or intra-articular steroid injection and physical therapy. Further trials are needed to gain a more comprehensive appreciation of the effects of intra-articular HA injections on patients at different stages of AC and with varying levels of pain and immobility.

Physiological mechanism of HA

HA, or hyaluronate, a large glycosaminoglycan, a high-molecular-weight biodegradable polymer, occurs naturally in large amounts in the extracellular matrices of soft connective tissue and synovial fluid and is a biomechanical and functional element of the articular cartilage.

HA contributes to physiological processes such as cell growth and movement in almost all tissue. It is used to treat osteoarthritis of the knee joint. Recent studies report that HA protects cartilage both through viscosupplementation and by suppressing the production of osteoarthritic-associated cytokines such as IL-1β, IL-6, and TNF-α37, and of ECM-degrading proteins such as MMP-1, MMP-3, MMP-13, and ADAMTS439 in human chondrocytes. Kataoka et al. ¹³ found that HA reduced MMP output after IL-1β stimulation in arthritis patients. In AC, in vitro trials have indicated that HA can decrease the breakdown of matrix components and stimulate proteoglycan synthesis. HA can stimulate vascular endothelial growth factor and type IV collagen, resulting in the acceleration of tendon healing. ¹³ HA at different concentrations significantly and dose-dependently reduced the levels of pro-inflammatory cytokines appearing in subacromial synovial fibroblasts in AC patients. ¹⁴ By growing the glenohumeral synovial/capsular fibroblasts from these synovial/capsular specimens, the researchers were able to examine the effect of HA on cell growth and messenger RNA (mRNA) output of the adhesion-related procollagens α1 (I) and α1 (III). Dependent on dose levels, HA significantly slowed cell proliferation and output of the mRNA of these procollagens in glenohumeral synovial/capsular fibroblasts for 24 h after stimulation, and inhibited the mRNA output of TGF-β and PDGF-A and PDGF–B. Another study that HA3000 significantly drew out the mRNA and protein output of MMP-1 and MMP-3 in IL-1β-stimulated tenocytes. ¹⁵ The ability of HA to decrease pro-inflammatory cytokine levels and to slow down the growth of adhesion-related cytokines may account for the clinical improvement after HA injection.

Some studies have suggested that the way HA passes through cellular structure in the tissues may be linked to CD44, which is the principal receiver of HA and has a number of physiological and pathological functions. The introduction of HA can affect cells by interacting with the family of CD44 receptors. There is some evidence that the interaction between HA and CD44 inhibits the output of pro-inflammatory cytokine mRNAs and COX-2/prostaglandin E2 (PGE2) in IL-1-stimulated subacromial synovial fibroblasts and suppresses the production of MMP-1, MMP-3, and MMP-13 in IL-1β-stimulated chondrocytes. ¹⁵ When Wu et al. pre-treated tenocytes with OX-50 to inhibit CD44 output, the inhibitory effects of HA3000 on IL-1β-stimulated MMP-1 and MMP-3 expression were significantly reversed. However, when the cells were pre-treated with inhibitor OS/37, blocking did not prevent the inhibitory effects of HA on the mRNA expression of adhesion-related collagens and cytokines. ¹⁴ CD44 may be the crucial factor in the pathway of HA’s effects on AC. More research is needed to elucidate the precise mechanisms of the intracellular signaling pathways involved.

Conclusion

At present, AC of the shoulder defies a definite clinical solution. There is a general acceptance of Neviaser’s definition and description of the pathological process of AC. It may be possible to target treatment better by understanding how the stages of the condition are linked to its fundamental pathophysiology. There is molecular evidence that the pathophysiological process of AC involves inflammation and damage to connective tissue. However, there is still no agreement on how this condition evolves and what can be done to treat it effectively. More research must be undertaken to clarify what actually takes place within the tissue structures. HA is a potential candidate for treatment, which works by decreasing the levels of pro-inflammatory cytokines and by slowing down the growth of adhesion-related cytokines, and CD44 may act as the crucial factor in the intracellular signaling pathways. However, the mechanisms by which it works are still unclear. More research and trials are needed to show at which stage of the condition HA treatment is most effective, whether it needs to be combined with other forms of medication and how it affects the intracellular signaling pathways.