Although animal studies have suggested that corticosteroid injections may have damaging effects on articular cartilage, human studies have not shown similar results [11, 12]. Intra-articular corticosteroid injections have an excellent safety record, and its safe use is supported by a large body of clinical data. The American College of Rheumatology has endorsed corticosteroid injections as safe and effective when administered by experienced physicians [2]. Despite their relative safe use, known side effects exist. Table 3.2 lists common and uncommon local adverse effects of intra-articular corticosteroids.

The most common reported side effects following corticosteroid injections are post-injection flair, facial flushing, and cutaneous atrophy [13].

Pain in the injected joint or at the site of injection can occur within the first 24 h after injection in up to 10 % of patients. Localized tissue damage resulting from needle puncture may be a partial cause for injection site reactions. More often, though, it is the crystalline structure of the particular corticosteroid agent used that results in a localized synovitis [14].

Facial flushing occurs within a few hours post injection in up to 15 % of patients and is particularly common in women. Although benign, symptoms may linger for up to 3–4 days [15].

Skin or fat atrophy following corticosteroid injections usually develops within 1–4 months. It occurs more commonly in patient with juvenile idiopathic arthritis and is reported in up to 8 % of injections [16]. Overall it is thought to occur at a rate of less than 1 % and may be accompanied with depigmentation of the skin [13]. The atrophy and depigmentation is due to the leakage of the injected steroids into the skin and may improve over a few months. It is more likely to occur following the small-joint injections where the accuracy of injection is not guaranteed, the subcutaneous tissue is thinner, and a larger volume of corticosteroids are injected. Less soluble agents are also more likely to cause these cutaneous adverse effects [3, 4].

Capsular calcification is the most common local adverse reaction following an intra-articular corticosteroid injection and occurs in up to 25–50 % of patients but is rarely clinically significant [17]. Pericapsular or intracapsular calcifications are noted within 2 months to 1 year following the injection and are usually asymptomatic. The location of the calcification is related to the site of the needle injection and is composed of hydroxyapatite [4].

Ruptured tendons occur uncommonly and have been reported in patients following intra-articular corticosteroid injections that have been placed directly within a tendon and may occur following a single injection in nearly 25 % of reported cases. Achilles tendon ruptures make up 50 % of reported cases followed by patellar tendon and biceps tendon ruptures in 19 and 8 % of cases respectively [18]. Great care should be used to avoid inadvertent injections within tendons.

The risk of causing a joint infection is one of the greatest concerns with the use of intra-articular corticosteroid injections, but it is one of the least common reported side effects. Pal B and Morris J reported the perceived risks of joint infection following intra-articular corticosteroid injections between 1 in 1,000 and 1 in 25,000 among rheumatologists surveyed [19]. The reported incidences in another study following knee injections ranged from 1 in 3,000 to 1 in 50,000 [20].

Systemic side effects on intra-articular injections are generally milder than with oral or intravenous formulations and often of unclear significance when present. Reported systemic side effects of systemic glucocorticoids include weight gain and fat redistribution, osteoporosis and fracture, osteonecrosis, ocular complications, hyperglycemia and diabetes, cardiovascular effects, infection, gastrointestinal complications, steroid-induced myopathy, hypothalamic-pituitary-adrenal axis suppression, and psychiatric complications. Table 3.3 lists drug-referenced side effects of glucocorticoids

Suppression of the hypothalamic-pituitary-adrenal axis following intra-articular injections is well-documented and usually mild and transient [3]. An average of 21.5 % reduction in serum cortisol levels returning to baseline after 72 h has been reported. Less commonly, prolonged HPA axis suppression lasting 5–7 weeks has also been reported [21].

Increased hepatic glucose synthesis and decreased insulin sensitivity have been shown to occur following corticosteroid therapy [22]. Intra-articular corticosteroid injections cause a transient increase in blood glucose levels; however no changes in fasting or predinner blood glucose readings were identified over a 2-week time period in a report of diabetic patients who received a methylprednisolone acetate injection for rheumatic complaints [23].

Steroid-induced myopathy is a known consequence of corticosteroid therapy and is more common with fluorinated corticosteroids (triamcinolone, dexamethasone) than with the non-fluorinated corticosteroids (hydrocortisone and methylprednisolone). It has not been reported following intra-articular injections [24].

Osteonecrosis occurs in 5–40 % of patients treated with oral glucocorticoids with increased incidence at higher doses and longer duration. It is reported rarely in oral doses less than 20 mg/day and has been reported following multiple joint injections within days to months.

Osteoporosis is a known side effect of systemic glucocorticoid use. Observational studies report the development of fracture-related bone loss in as high as 40 % of patients with systemic glucocorticoid use [25]. In contrast, no net effect on bone resorption and only a transient effect on bone formation were found in a study following single intra-articular triamcinolone acetonide injections [26]. A theoretic benefit of increased mobility following intra-articular injections may also counteract osteoporotic effects.

The rationale for using corticosteroids in the treatment of arthritis is to suppress inflammation, suppress inflammatory flares, and disrupt the inflammatory damage-repair-damage cycle. While the clinical efficacy of intra-articular glucocorticoids is well described, there is very limited data on the effects on human synovial tissue in inflammatory arthritis. The systemic effects of glucocorticoids are well described and mediated by receptor antagonism of nuclear factor kappa beta (NF-kB). NF-kB antagonism results in decreased transcription of several genes involved in the immune inflammatory response. It has also been shown to destabilize mRNA. These effects result in a dramatic reduction in cytokine production yielding pleotropic results to include effects on endothelial cells, cell migration, monocytes, macrophages, neutrophils, eosinophils, lymphocytes, fibroblasts, bone, cartilage, muscle, and the synovial tissue [27]. An in vivo study examining the synovial biopsy tissue of 31 patients with inflammatory arthritis, mostly RA, sheds some light on what happens in the synovium. Synovial protein expression of tumor necrosis factor (TNF), interleukin (IL)-1 beta, extranuclear high-mobility group box protein (HMGB)-1, vascular endothelial growth factor (VEGF), and ICAM-1 was reduced but without significant effects on vascularity [28]. Notably, macrophage infiltration and proinflammatory endothelial cytokine expression were not reduced, possibly explaining the transient nature of improvements with intra-articular corticosteroids. In this study, all patients demonstrated clinical improvements, although it should be noted that many patients were on other DMARDs or systemic corticosteroids [28]. More recently, synovial citrullinated protein has been correlated with the degree of local inflammation. Treatment with intra-articular corticosteroids has been shown to alter the expression of the synovial citrullinated proteins in the inflamed joint. This downregulation has been correlated with glucocorticoid effects on a specific protein, peptidylarginine deiminase 4 (PAD4). PAD4 activation, induction, and signaling pathway is dependent on NF-kB suggesting glucocorticoids may affect citrullination through PAD4 downregulation in an NF-kB-dependent pathway [29, 30]. A summary of the described cellular effects following glucocorticoids is listed in Table 3.4.