Statistics from Altmetric.com
The ugly sister
In this issue of the BJO (p 1380), Ebner and colleagues present data from their pilot study of orbital steroids in Graves’ orbitopathy. Even though the patient population is small, and the treatment groups not entirely comparable, the study does provide some evidence that steroid injections used to treat Graves’ orbitopathy may be safe and efficacious.
Orbital steroid injections for Graves’ orbitopathy and other inflammatory disease of the orbit have had a somewhat chequered past. During my training under Norman Shorr, MD, at the Jules Stein Eye Institute, UCLA, I observed frequent use of orbital steroid injections to treat orbital inflammatory disease including Graves, and I also learned about steroid injections from William Stewart, MD, in San Francisco. As I subsequently travelled to different institutions as a visiting speaker, I was surprised therefore to find that in many programmes orbital steroids were never used, either because of concerns about safety or concerns about efficacy. Compared to oral steroids, they were considered the ugly sister.
Certainly, corticosteroids are a potentially dangerous medicine with a long track record of proved ability to cause significant complications. In addition to the systemic complications of steroids, injection around the eye also poses the risk of local complications including globe perforation,1–3 intractable elevated intraocular pressure,4,5 conjunctival or corneoscleral melting,6–8 vascular occlusion from embolisation or pressure induced optic nerve compression,9–16 proptosis or fat atrophy,17–19 depigmentation, and granuloma related to the methyl cellulose vehicle of the depot injection. However, I suggest that reluctance to use orbital steroid injections derives its legacy from complications of retrobulbar steroid injections used to treat intraocular processes. Retrobulbar or sub-Tenon’s injections are given with a long needle, and the medication is very close to or in direct contact with the globe.
The technique for orbital injections is substantially different from retrobulbar injection. For orbital injections, the medication is placed just inside the septum near the orbital rim. A half inch needle is used (I prefer a 27 gauge needle to allow more easy passage of the particulate injection) and the risk of perforating the globe is remote. Also, the small short needle is well tolerated by patients: by using distraction techniques and a gentle touch, injections can be given almost painlessly. Perhaps because the medication is not in contact with the globe, severe elevation of intraocular pressure is not typically seen (although I would be careful in recommending orbital steroid injections in patients with significant glaucoma). Triamcinolone acetonide (Kenalog) is manufactured in a 40 mg/ml concentration, so no more than 1 ml of fluid is needed. Particularly in the anterior orbital location, with slow injection, I have not observed and would not expect a 1 ml injection to create a vision threatening increase in orbital pressure. Slow injection using the small gauge needle is also important in order to minimise the risk of intravascular injection forcing retrograde flow of particulate matter into the retinal circulation.12,20
Efficacy is a difficult thing to measure in Graves’ orbitopathy; the heterogeneity of the disease and the tendency for spontaneous improvement, create a requirement for carefully controlled studies with adequate numbers of patients in order to assess and compare treatment options. I am not aware of any studies of steroid injections, including the pilot study published here by Ebner et al, that definitively answer the question of efficacy of steroid injections relative to disease outcome.21
On the other hand, no clinician who cares for patients with Graves’ orbitopathy would likely dispute the efficacy of corticosteroids in ameliorating the acute inflammatory features of the disease. Corticosteroids have been a mainstay of treatment, and often provide rapid improvement in symptoms and in clinical evidence of inflammation. Whether the temporary improvement observed with corticosteroid therapy affects the final outcome of the disease is a fascinating and unanswered question, but at least the temporary improvement offered by corticosteroids often allows these patients to resume daily activities.
We should not require studies of injected steroids to show proof that they fundamentally change the disease. The pertinent question is whether steroid injections can achieve the improvement in symptoms that we observe with oral corticosteroids, without an unacceptable rate of local complications, and with decreased systemic side effects. The anecdotal experience of those who use this delivery mechanism suggests that injected steroids do provide symptom relief. Although there is some systemic absorption,21,22 pharmacological studies of intra-articular injections suggest good compartmentalisation22 and anecdotal experience, as well as Ebner’s study, suggest decreased systemic side effects compared to oral steroids. With attention to careful technique, I believe they can be safely used (although patients have to accept the rare risks of steroid injections which include blindness). In the absence of large studies to accurately determine the true complication rate, practitioners should be encouraged to report major complications to the medical community.
Important technical points to improve the safety of injection include the use of a short half inch 27 gauge needle, slow injection, injection site in the anterior orbit just posterior to the orbital septum (fig 1), and limitation of injected volume to 1 ml or less. The depot injection lasts for about 30 days (without prompting, patients often volunteer that the improvement in symptoms wears off after a month) so that I personally use a dosing schedule of one injection per month, repeated as needed. I have injected as many as 13 aliquots in patients with recurring orbital myositis, but on average, I find that patients with Graves’ orbitopathy need treatment for 3–6 months. My anecdotal experience is that a 40 mg injection provides an effect equivalent to 20 mg daily of oral prednisone, so that patients who can be controlled at that oral dose are good candidates. Systemic side effects are minimal, or at least reduced, compared with oral steroids. I am sure that there is some placebo effect associated with the dramatic experience of an injection next to the eye, which works in our favour as well.
Our patients with Graves’ orbitopathy have a tough battle. They have a frustrating disease that still awaits a cure. The more options that we have for medical and surgical symptomatic treatment, the better we can help them through the stages of their disease, and get them back to a comfortable, productive life. Orbital steroid injections seem to deliver a moderate dose of depot corticosteroid to the orbit with an acceptable local complication rate and with decreased systemic toxicity relative to oral corticosteroids.
Support provided by the Dr Bernard and Jenny Nelson Family Trust.
I thank Michael Roberts, MD, and Guy Ben Simon, MD, who provided valuable assistance in the preparation of this manuscript.
Note in Proof
The ugly sister