Design and Methods: Ten patients were selected for treatment in an open study. Ophthalmological assessment and orbital magnetic resonance imaging (MRI) was performed to determine disease activity, severity and outcome of immunosuppression. Patient sera was tested for detection of antibodies (Ab) reactive with pig eye muscle membrane (PEMM) proteins in immunoblotting.
Results: Clinical activity score (ranged 0-7) fell significantly in all patients within the first week of treatment from (mean ± SD) 5.6 ± 0.84 to 2.6 ± 0.84 (p < 0.01) and remained stable 12-36 months later. This was accompanied by a reduction in proptosis, upper lid retraction and lagophthalmos with healing of keratitis. Intraocular pressure was normalized and extraocular muscle shrinkage could be observed in MRI. Two patients required additional optic nerve decompression. Titers of antibodies against PEMM proteins decreased significantly but remained detectable during follow up. Pretibial myxedema disappeared in a patient in whom hyperthyroidism relapsed by the end of treatment. Side effects were minimal and transient.
Conclusions: Treatment with i.v. MP pulses followed by oral prednisolone is effective, safe and well tolerated for control of the TAO. It has to be determined if changes in PEMM Ab titers influence prognosis and if they could be useful for prediction of TAO relapses
Thyroid-associated ophthalmopathy (TAO) is a disfiguring and disabling disorder that remains a challenge for patient management and research. Accumulation of mononuclear cells (mainly activated T lymphocytes, but also B cells and macrophages) within the extraocular muscles in patients with untreated TAO is accepted as evidence of the autoimmune nature of the disease1. Actions of T cells reacting with an as yet unknown orbital/thyroid cross-reactive antigen, along with macrophages, which release cytokines that active the orbital fibroblasts leading to increased production of glycosaminoglycans, interstitial edema and, finally fibrosis may explain the evolution of the characteristic features of TAO2. The identity of primary autoantigens for the orbital immune response as well as the mechanisms for tissue damage remains unclear and controversial. Notions about the pathogenesis of TAO have focused upon either the extraocular muscles (EM) or the orbital fibroblasts as best candidates for the immune target and the nature and role of various effector cells is also debated3. In addition, putative links between orbit and thyroid antigens have not been completely elucidated4. Antibodies (Abs) against eye muscle membrane proteins, with mol wt 55, 64 and 95 kDa have been found to be associated with TAO by some researchers5-8 but not by others9. Whether these Abs play a pathogenetic role in the ophthalmopathy or reflect the tissue damage as a secondary event is unclear. The likely role of T cells in the pathogenesis of TAO has not been studied extensively in view of restricted access to EM from patients with recent and active disease. The 64-kDa protein was also demonstrated expressed in thyroid10 and the major clinical utility of measuring such antibodies is related to its predictive value for the development of ophthalmopathy11,12. More recently, Kubota et al have sequenced and identified the 55 and 64 kDa proteins13. While TSH-receptor antibodies (TSH-R Ab) play a leading role in Graves' hyperthyroidism, their possible participation in the pathogenesis of extra-thyroid manifestations of the disease has not been proven14 and appears unlikely15. Since specific immunosuppression is not available, oral glucocorticoids and orbital irradiation are most widely used to manage active TAO, but their side effects and/or effectiveness limit practice16,17. Intravenous pulses of high doses of methylprednisolone (i.v. MP), as an alternative treatment appears to improve the outcome of TAO in recent open studies, allowing a rapid control of inflammatory manifestations with less side effects22-25. We report now the results of our experience bared upon orbital changes, clinical effects and tolerance, and variation observed in some indicators of the auto-immune response associated with treatment of 10 patients with severe TAO with high dose i.v. MP pulses and oral prednisolone.
PATIENTS AND METHODS
Orbital/ophthalmologic changes assessment
Disease activity and objective ocular changes were assessed according to recommendations of an International TAO Workshop18. Clinical Activity Score (CAS) graded 0-7, was determined for disease activity from the findings of: spontaneous retrobulbar pain, pain on eye movement, eyelid edema, eyelid erythema, conjunctival injection, chemosis and swelling of the caruncle; each scoring one point.
Eyelid position measurements, namely lid lag, retraction and lagophthalmos were taken in relation to the corneoscleral limbus (mm). Proptosis (mm) was measured with a Hertel exophthalmometer and MRI. EM thickness was assessed by T1-weighted image in MRI; signal intensities of EM were studied by short time inversion recovery (STIR) image for differentiation of edematous changes from fibrosis as recommended previously by Hiromatsu et al19. Intraocular pressure measurements were taken at primary sigh position and at 15° on upward or attempted upward gaze; differential intraocular pressure was also considered for indirect evaluation of EM enlargement (normal ¾ 3 mmHg). EM function was studied by the cover test and Lancaster test. Corneal status was visualized using fluorescein test and slit lamp observation. Visual acuity, visual fields, chromatic vision, and funduscopy were used to document optic nerve status and function. Disease severity was defined according the "NO SPECS" classification20.
Self-patient assessment of the eye condition was also included, using a scale of 0 to10 (worst to best) in respect to appearance, eye discomfort, diplopia and visual acuity.
Patients
Patients were selected for treatment when they presented progressive symptoms and external signs of inflammation (CAS >= 4), progressive proptosis (>= 4 mm), EM swelling, corneal damage and/or optic nerve impairment, with no previous treatments for the eye disorder. The following clinical entities were considered contraindications for corticosteroids: severe psychiatric disorders, active viral infections, active tuberculosis, cardiac arrhythmia, congestive heart failure, anemia, atopia and active peptic ulcer. Ten patients with severe and active TAO were selected for the study: 8 female and 2 male, aged 34-59 yr. (mean 49.2 yr.), with a mean duration of ophthalmopathy of 11.1 months. All patients suffered from associated Graves' hyperthyroidism and were maintained in an euthyroid state throughout the period of observation, which was at least three months prior treatment, and during a follow up of 12-36 months. Seven patients were treated with anti-thyroid drugs, and the other 3 patients were under thyroxine replacement for hypothyroidism following radioiodine treatment. Two out of the ten patients had type 1 diabetes mellitus, which was not considered a contraindication for corticosteroid therapy; one patient had clinically evident pre-tibial myxedema (table 1) and eight were heavy smokers. Optic nerve compression was identified in two patients before steroid therapy in whom visual acuity was severely reduced, along with visual fields and chromatic vision impairment, while papiledema could be seen in funduscopy. Absence of single binocular vision was found in all patients before treatment in whom orbital MRI showed EM enlargement, with signal intensities suggestive of edema, in mainly medial and the inferior rectum muscles.
The TAO Committee of our Hospital evaluated all cases for treatment. The Hospital's Ethics Committee approved the study and informed consent was obtained from all patients.
Treatment protocol
Intravenous 6-methylprednisolone-sodium hemisuccinate (MP) 500 mg in 250 cc isotonic saline solution was infused over 60 min on each of 3 consecutive days. MP pulses were followed by oral prednisolone, 40 mg daily the first week, and then tapered down by 10 mg weekly during the following four weeks up to a daily dosage of 10 mg. A second course of therapy was repeated one month later in all patients who had a good response to the first treatment. Ranitidine (150 mg/12 hr) and potassium replacement were added to patients with history of gastritis or peptic ulcer and when ionogram showed hypokalemia respectively. Clinical (complete physical examination, blood tension and weight control) and laboratory assessment (glucose and potassium levels) was performed in each day of methylprednisolone i.v. infusion, at the first month and at the end of treatment (second month). After prednisolone withdrawal, hydrocortisone 20 mg/day was maintained until recovering of adrenal function (basal cortisol >= 275 nmol/l, cortisol >= 600 nmol/l at 30 min post 250 µ g Synacthen® i.v, and a difference >= 220 nm/l between basal and 30 min).
SDS-PAGE and Western blotting
Fresh eye muscle was obtained from normal pigs immediately after sacrifice, separated from adipose and connective tissues, minced and homogenized in a Polytron mechanical blender. The homogenate was centrifuged at 3,000 rpm for 20 min to remove cell debris, nuclei and fat. The supernatant was centrifuged at 100,000 g for 60 min, then discarded, and the membrane pellet washed
three times and finally reconstituted in phosphate-buffered saline (PBS). Pig eye muscle membrane (PEMM) proteins were separated by standard Laemmli's sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using 8% separating gel and 4% stacking gel, run at 110 V under reducing conditions with mercaptoethanol. Proteins were then electroeluted to polyvynilidendifluoride (PVDF) papers at 100 V for 1h. After blocking with 10% polyvinyl pyrrolidone in Tris-buffered saline (TBS) the strips were incubated overnight at 4 ºC with patient sera (obtained before, during and after steroid treatment). Serum dilutions ranging from 1:10 to 1:1600, the concentration of antibody being expressed as a titer, i.e., the greatest serum dilution when a band at 64 kDa was just visible, according Miller et al11. The PVDF strips were then washed and incubated with alkaline phosphatase-conjugated goat antihuman IgG antiserum (* chain specific) diluted 1:500 in TBS for 2 h at room temperature and developed with 5-bromo-4-chloro-3-indolyl phosphate toluidine and p-nitroblue tetrazolium chloride.
Statistical analysis
The Wilcoxon matched pairs signed rank test was used to analyze the observed objective ocular changes, self-patient assessment Score and PEMM Ab titers. The Friedman test of significance was used for repeated measured analysis of variance for CAS values and PEMM Ab titers during treatment (0 to 8 weeks) and throughout the first year of follow up. Spearman's test was employed to analyze correlation among changes in Ab levels and CAS.
RESULTS
Ocular/orbital changes:
Mean (± SD) values for CAS fell significantly already the first week of therapy from 5.6 ± 0.84 to 2.6 ± 0.84 (Wilcoxon test, p < 0.01) (fig. 1), remained stable (2.6 ± 1.17) by the end of treatment, and fell to low levels (0.5 ± 0.70) during follow up 12-36 months later. Hertel's and MRI exophthalmometry measurements showed that the amount of proptosis was significantly reduced (table 2). EM shrinkage was also observed, as determined by MRI in 9 out of 20 (45%) orbits (MRI orbital changes of case 9 are shown in fig. 2). Overall, the differential intraocular pressure was not significantly changed (table 2), although initially abnormal values were shown in only five out of the twenty explored eyes; in all of these, values returned to normal by the end of treatment. Visual acuity, when accompanied by papiledema, was not substantially modified by steroid treatment in the two affected patients; in both patients optic nerve decompression was performed after (patient 6) and during (patient 10) steroid therapy, respectively. Following transnasal endoscopic surgical decompression (medial wall removal), visual acuity was fully recovered in both patients. Upper lid retraction was significantly reduced from (median, range) 2 mm (0-6) to 0 mm (0-4) [p < 0.001] after treatment. Similar improvement was achieved in lagophthalmos from (median, range) 1 mm (0-4) to 0 mm (0-2) by the end of treatment (table 2), allowing healing of keratitis in the eight previously affected patients.
| 
|
Single binocular vision was completely restored in 6 out of 10 patients after treatment; the remaining 4 patients with persistent diplopia of different degrees required surgical correction in three of them and prisms in the other one. Pretibial myxedema (in patient 3) completely disappeared after the two months of corticosteroid therapy. Self-patient assessment also showed the achievement of favorable changes in appearance, eye discomfort, diplopia and visual acuity at the end of treatment (table 3).
PEMM antibodies
Reactivity against a 64 kDa PEMM protein was detected on immunoblotting with patients sera before therapy, and their titers decreased significantly after treatment in all patients, being detectable during follow up in all cases (fig. 3) (PEMM Abs changes in case 10 is illustrated in fig. 4). Spearman's coefficients showed no overall correlation between changes in PEMM antibody titers and changes observed in CAS.
Adverse effects and other observations
Hypokalemia was detected in 4 patients during treatment, 1 patient had a slight elevation of arterial blood tension, mild weight gain (around 2 kg) was registered in 2 patients, rounded face was observed in 2 patients, and transient hyperglycemia was detected in a non diabetic patient. One of the diabetic patients, who was simultaneously treated with macrolide antibiotics for pharyngitis, required an increase in her insulin dose three times during steroid treatment. In this patient, who also developed mild erosive antritis as well as hypokalemia and rounded face could be observed. All side effects completely disappeared after withdrawal of corticosteroids in all patients. One patient (n.o 3) had a relapse of her Graves' hyperthyroidism at the end of the 2nd month of steroid course, being previously hypothyroid on thyroxine replacement for one year after radioiodine treatment; that relapse was associated with improvement in her ophthalmopathy and disappearance of signs of pretibial myxedema. TSH-receptor reactive antibodies were measured in that patient, as TSH-receptor binding inhibitory immunoglobulins (TBII) in a standard radioreceptor binding assay and elevated levels of 35 U/L (upper limit of normal = 9 U/L) were detected at the same time. Upper limbic conjunctivitis (Theodore conjunctivitis) rearmed in patient n.o 4 when prednisolone was withdrawn, increasing the mean value of CAS.
DISCUSSION
We observed a rapid clinical response to high dose i.v. methylprednisolone pulse therapy, as regression of inflammatory signs and symptoms was noticed after 1-7 days of treatment in all patients, and in most cases within the first 24 hours. While generally recommended18 but not still widely used, the CAS is considered by our group as an easy and objective tool for evaluation of disease activity in clinical practice, not only for better patient selection for treatment, but also in respect to the outcome of immunosuppression21. Our data show favorable changes in many other parameters of ophthalmopathy in 10 patients with severe disease. Proptosis and eye muscle thickness was reduced and eyelid retraction was adequately restored to enable healing of keratitis by avoiding corneal exposure. Even though visual acuity was not substantially modified in the two affected patients, the treatment allowed better conditions for surgical decompression. Kendall-Taylor et al22 reported similar limitations of i.v. MP-P therapy to achieve a rapid restoration of visual acuity when it was accompanied by papiledema. All favorable changes were persistent during long-term follow up of 12-36 months. Despite no definite agreement about the proper dose, frequency and duration of iv MP pulses for the treatment of TAO, it is well accepted that its rapid clinical effect is specially useful for severe disease, providing better response and fewer adverse effects than high oral doses of prednisone22-25, although a randomized controlled trial was not yet performed to confirm this. Being the most potent anti-inflammatory drugs, glucocorticoid therapeutic strategies are empirical and based on symptom control, correction of function and avoidance of distressful adverse effects. Side effects could be related to age, sex, concomitant diseases, interactions with other drugs, as well as genetic factors that influence glucocorticoid metabolism. Methylprednisolone has a greater volume of tissue distribution and is more susceptible to drug interactions than prednisone and its active metabolite prednisolone26. This fact may explain, in part, the better outcome and rapid response observed after the MP pulses when compared with other regimes of steroid treatment, and the slight differences of susceptibility to side effects between individual27. In our patients, only transient and mild side events were observed and in no case they represented a reason to stop the treatment. In case 2, an elderly diabetic subject who was simultaneously treated with macrolide antibiotics, the development of those side effects was observed, probably due to potential interaction with methyl-prednisolone pharmacokinetics26. As suggested by Hill et al28 the dosage regimen should be individualized for each patient with the goal of obtaining maximal benefit with minimal risk. PEMM antibodies fell during treatment with parallel improvement in signs of clinical ophthalmopathy, although remaining positive at the end of treatment in all cases. It is well known its predictive value for TAO development11,12. These antibodies determination may be a good marker for TAO outcome after treatment, and the follow up of this group will provide information about their predictive value for TAO relapse. Relapse of hyperthyroidism with elevated levels of TSH-receptors antibodies while improving eye condition and pretibial myxedema in one patient suggests that those antibodies does not play a role in the pathogenesis of extra-thyroid manifestations in Graves' disease.
In conclusion, i.v. MP pulses, followed by oral prednisolone, is an effective, safe and well tolerated therapeutic alternative for TAO control, particularly for patients with active and severe disease, although larger series are necessary to confirm our observations. We suggest that rapid beneficial effects on TAO may be preferably exerted by an anti-inflammatory mechanism that seldom cures the eye disease but change its natural course reducing sequelae. While it needs to be proven that changes in PEMM Abs titers influence prognosis, these Abs may be helpful for prediction of TAO relapses.
ACKNOWLEDGMENTS
SW was supported by a grant from LAIR Foundation, Madrid, Spain.