Adjuvant Treatment of Graves´ Ophthalmopathy With NSAID (aGO Study)

Adjuvant treatment of Graves´ ophthalmopathy with NSAID (aGO study)

1. Background. When performing MRI / CT / ultrasound of the orbital room 98 % of the patients with Graves´ thyrotoxicosis exhibit changes of ophthalmopathy without clinical symptoms of endocrine ophthalmopathy (Burch 1993). Thus, almost all patients with Graves´ disease are at risk to develop clinical ophthalmopathy. Tobacco exposition and treatment with radioiodine ar known factors that trigger ophthalmopathy but other factors are poorly defined. However it is well known that both environmental and genetic factors is of importance because monozygotic twins show a concordance of 20 - 50 %. One interesting observation is a patient with inactive and stable Graves´ ophthalmopathy who developed clinical ophthalmopathy when treated with pioglitazone due to type 2 diabetes (Starkey 2003). One pathogenic mechanism in Graves´ ophthalmopathy is increased orbital adipogenesis and glitazones are known to increase the volume of subcutaneous adipose tissue. Orbital fibroblasts from patients with ophthalmopathy have been shown to differentiate to adipocytes in response to rosiglitazone (Valyasevi 2002). Glitazones are PPAR-gamma agonists and it could be of interest to study effects of PPARgamma antagonists in endocrine ophthalmopathy. Diclofenac has been shown to interact with PPAR-gamma in physiological concentrations and to antagonise PPAR-gamma mediated effects like adipogenesis of the preadipocyte cell line 3T3-L1(Adamsson 2002, Vondrichova 2007). In addition diclofenac is a well-known inhibitor of cyclooxygenases with effects on the synthesis of prostaglandins in for example fibroblasts, lymphocytes and monocytes. The natural ligand for PPAR-gamma is prostaglandin J (Forman 1995). To conclude, NSAID like diclofenac may affect both synthesis of prostaglandins and concomittantly antagonise the effects of the natural ligand to PPAR-gamma, prostaglandin J2. We have recently demonstrated upregulation of immediate early genes, including COX-2, with important functions in adipogenesis in patients with severe ophthalmopathy (Lantz 2005). Previously there is only one study published on treatment of ophthalmopathy with NSAID ,indometacin (Amemia 1982). Although only 7 patients were studied there were effects on soft tissue symptoms, eye protrusion and eye muscle symptoms. Due to the knowledge of retrobulbar morphological changes in patients with Graves´ thyrotoxicosis without clinical ophthalmopathy it may be of importance to interact as early as possible. We therefore plan to start a randomised single-blind study where patients will receive diclofenac for 12 months in parallel with the standard treatment for thyrotoxicosis.
2. Trial design The trial is a multi-center (Malmö and Stockholm), controlled, single-blind design and the subjects are randomised to thyrostatics with and without diclofenac by a body independent from the study The total duration of the study is 24 months with 12 months treatment with or without diclofenac in addition to regular treatment for thyrotoxicosis. The follow up period will be 12 months. Decision of the main treatment of thyrotoxicosis is based on clinical terms according to routines in Malmö and Stockholm and is given according to the protocol of TT96. Patients are randomised for adjuvant therapy with or without diclofenac after choice of main treatment for thyrotoxicosis. 2.1 Monitoring The study will be monitored by a science nurse with formal education in monitoring clinical trials.
3. Objectives

The primary objective

To find out if diclofenac can prevent development of ophthalmopathy in patients with Graves´ disease.

The secondary objectives are:

1. Determine the activity at diagnosis of ophthalmopathy
2. Determine the time from thyrotoxicosis to ophthalmopathy
3. Determine the frequency of patients with corticosteroid requiring ophthalmopathy 3.1 Endpoints The objectives of the trial will be examined by the following endpoints: Efficacy Primary endpoint: The frequency of ophthalmopathy after 24 months as judged by the following clinical signs:. Optic nerve dysfunction 0. No 1. Yes Eye-lid edema 0. No 1. Yes Chemosis 0. No 1. Yes Conjunctival injection 0. No 1. Yes Exophthalmos 0. No 1. Yes Hertel - base right left Eye muscle dysfunction 0. No 1. Yes Corneal ulcers 0. No 1. Yes Summa: Ophthalmopathy is present if the patient has one sign or more. Secondary endpoints:
4. Activity of ophthalmopathy as judged by clinical activity score (CAS) Spontaneous retrobulbar pain 0. No 1. Yes Painful eye-movements 0. No 1. Yes Eye-lid erythema 0. No 1. Yes Conjunctival injection 0. No 1. Yes Chemosis 0. No 1. Yes Swollenness of caruncula 0. No 1. Yes Eye-lid edema or swollenness 0. No 1. Yes Sum: II. The time from thyrotoxicosis to ophthalmopathy. III. The frequency of corticosteroid requiring ophthalmopathy. Criteria for start of steroid treatment are: 1. Risk of corneal ulcers with or without exophthalmos 2. Double vision within 30 degrees 3. Optic nerve dysfunction. 3.2 Treatment of subjects All patients that fulfill the inclusion criteria and accept participation in the study will after decision of main treatment for thyrotoxicosis be randomised to treatment with or without diclofenac. Patients will receive diclofenac 50 mg 1x2 for 12 months in addition to standard treatment of thyrotoxicosis according to the following routines. Thyrostatics Patients will receive p.o. methimazole (Thacapzol) 5 mg 3x2 supplemented with L-thyroxine (Euthyrox) after 14 days, initially for the first two weeks 0.5x1, and thereafter 1x1. If the patient react with allergic symptoms to methimazole, treatment is changed to propylthiouracil (Tiotil) 50 mg 3x3. Adjustment of L-thyroxine is done with help of thyroid laboratory parameters (TSH, fT4) and titration after clinical response. The dose of thyrostatics is not reduced and if necessary on demand the dose is increased. If necessary, a beta-blocker is used, primarily propranolol (Inderal) 40 mg 1x1-3 or metoprolol (Seloken) 50 mg 1x1-3 which can be increased depending of patients response. 3.3 Measurement of laboratory parameters All laboratory analyses will be performed at the department of Clinical Chemistry, Malmö University Hospital and at the department of Clinical Chemistry Karolinska University Hospital.
5. Rationale for treatment Adjuvant treatment already at diagnosis of thyrotoxicosis has been chosen in the aim to prevent clinical ophthalmopathy due to the fact that morphological changes of ophthalmopathy in the orbital room exist in Graves´ thyrotoxicosis in up to 98% of the patients without clinical symptoms. The long period of time for treatment with diclofenac is justified by the knowledge that the majority of patients at risk for developing ophthalmopathy exhibit clinical symptoms within 18 months.
6. Subjects 5.1 Number of patients A total of 150 statistically analysable patients (the total number of patients included will be 10% higher to compensate for loss of subjects) , with 75 in each treatment arm will be included in the study.
7. Visits - summary

S = Screening tests, TSH, fT4, fT3, Thyroid Receptor antibodies (TRAK), Tissue Peroxidase antibodies (TPO-ak), Hb, leukocytes, thrombocytes (trc), diff.-leukocytes, fasting B-glucose, HbA1c, Calcium, Albumin, Sodium, Potassium, Creatinine, ASAT, ALAT, ALP, GT, Bil, PK, APTT, U-iodine R = Routine tests, TSH, fT4, fT3, Hb, Creatinine R+ = Routine tests + ASAT, ALAT, ALP, GT, Bil, TRAK, trc, leukocytes, diff. R++ = Routine tests + + Fasting B-glucose, U-iodine B = Biobank samples according to GD2002 (serum, plasma, buffy coat) C = Clinical control - vital signs, physical examination, weight, eye status O = Ophthalmologist - eye status D = Dietary questionnaire

Time

-1 week S, B, C, O, D 0 week C 6 week R, C 3 months R+, B, C 6 months R, B, C 9 months R+, B, C 12 months R++, B, C, O, D 15 months R+, B, C 18 months R, B, C 21 months R, B, C 24 months R++, B, C, O, D

6.1 Screening

Subjects attend to the first screening visit one week before randomisation. Before screening the participants have been provided with written information of the trial. The subjects will be informed, orally and in writing of their responsibilities / rights during the trial as well as possible advantages / disadvantages of the trial. Subjects who wish to participate will be asked to sign and date an Informed Consent Form prior to any trial-related activity.

Subjects will be allocated a subject number and the following will be performed and recorded:

1. Assessment of inclusion and exclusion criteria
2. Demographic information
3. Thyroid history
4. General physical examination
5. Measurement of body weight
6. Vital signs
7. Blood sampling
8. Pregnancy test will be performed as per judgement of the investigator
9. Concomitant illness and medication.
10. Information and choice of the routine treatments for thyrotoxicosis
11. Actual eye status, performed by an ophthalmologist within 1 week
12. An appointment will be made for visit w0

6.2 Randomization visit, w0

Laboratory data and other information gathered at screening visit is reviewed and the final decision regarding participation in the trial is taken.

The subject will now be randomised to adjuvant treatment with or without diclofenac. The randomization procedure will be designed by an independent body.

1. Trial material

The trial is not performed in direct collaboration with the drug manufacturers.

1. Statistical considerations

The primary evaluation variable is the proportion of present or new cases of endocrine ophthalmopathy. The null-hypothesis is that there is no differences between the two treatment alternatives according to this proportion (H0:pi 1 = pi 2). The number of patients in each group when performing a two-side significance test on the 5% level for different demands on power and different values in the given proportions of the hypothesis is shown in a separate diagram. With 72 patients in each group there is 80 % chance (power) to get significant results if the real difference is 0.2. The hypothesis is tested with Mantel Haenzels test and further analyses performed with logistic regression which take into consideration the effect of prognostic factors.