Introduction and objective: Corticosteroids may cause tendinopathies, an adverse effect that is not well known and characterised, although it was initially described more than 40 years ago. This study was conducted in order to characterise the important aspects of this adverse reaction, such as the role played by routes of corticosteroid administration, therapy duration, comedication, cumulative corticosteroid dose and underlying disease.
Study design and methods: Published case reports of tendinopathies that were associated with corticosteroid use were identified by a comprehensive literature search using the databases MEDLINE, Pharm-line, EMBASE, ToxFile, Adis Inpharma, International Pharmaceutical Abstracts, Drug Information Fulltext and PASCAL. The reference lists of all pertinent articles were cross-referenced to retrieve additional cases. Spontaneous reports were requested from the Uppsala Monitoring Centre (the WHO Collaborating Centre for International Drug Monitoring). Information of published and spontaneous reports was analysed with regard to age, sex, underlying disease, individual corticosteroids, equivalent corticosteroid dose, latency time, cumulative dose, route of administration, comedication and type of tendinopathy.
Results: We included and analysed 73 published case reports and case series involving 133 patients and 191 spontaneous reports of corticosteroid-associated tendinopathies. The proportion of women with tendinopathies was 50.8% and 41.0%, respectively. The mean age (+/- standard deviation) of patients with a tendinopathy was 50 +/- 17 years and 61 +/- 16 years, respectively. The predominant routes of administration were oral (33% of published cases and 47% of spontaneous cases) and intra-articular (35% of the published cases or parenteral (7% of the spontaneous cases). There were isolated cases of tendinopathy after inhaled or topical (i.e. ocular, cutaneous, nasal) corticosteroid use. Tendinopathies that were reported in the literature cases consisted mainly of tendon ruptures (93%), predominantly of the Achilles tendon. Of the spontaneous cases, 31% had a rupture and the remaining cases had a tendinitis or an unspecified tendon disorder.
Conclusion: Oral and parenteral applications, especially intra-articular use, were the most prevalent routes of administration in cases with corticosteroid-associated tendinopathies. However, topical application has also been rarely associated with tendinopathies. Future pharmacoepidemiological studies should further address this issue to quantify the risk of corticosteroid-associated tendinopathies.