Open Label Extension of a Clinical Trial of Intravitreal Triamcinolone for Diabetic Macular Oedema-TDMX Study

• Increase of ≥5 letters at the 5-year study visit on a LogMAR chart compared with (a) the initial baseline level and (b) the level at the 2-year study visit. [ Time Frame: 3 year extension, total 5 years study from baseline ] [ Designated as safety issue: No ]

  Changes from Baseline to 5 years: Improvement of ≥5 letters after 5 years was found in 14/33 (42%) eyes initially treated with triamcinolone compared with 11/34 (32%) eyes initially treated with placebo (zGEE=0.81, P=0.4).

  Changes from 2 to 5 years (open-label extension):Improvement of ≥5 letters of best-corrected visual acuity was found in 8/29 (28%) eyes initial-triamcinolone compared with 7/28 (25%) initial-placebo eyes (zGEE=0.20, P=0.8).

  
  
• Incidence of moderate or severe adverse events over the 3 years of the open-label extension [ Time Frame: 3 year extension study, total 5 year study from baseline ] [ Designated as safety issue: Yes ]
  The incidence of cataract surgery declined in the third year: 5/11 (45%) eyes from the initial-triamcinolone group that were phakic at the beginning of the 3rd year required cataract surgery.
  
  

• Change in macular thickness by OCT [ Time Frame: 3 year extension, total 5 year study from the baseline ] [ Designated as safety issue: No ]

  Changes from Baseline to 5 years: Foveal thickness had decreased by 30µm (95% confidence interval, -47 to 107µm) less in the initial-triamcinolone group than in the initial-placebo group at 5 years (zGEE=0.76, P=0.45).

  Changes from 2 to 5 years (open-label extension):Foveal thickness had actually increased slightly on average in the initial-triamcinolone group, but decreased in initial-placebo eyes. Overall it had decreased by 70µm (95% confidence interval, -1 to 140µm) more in the placebo group than in the treatment group between 2 and 5 years (zGEE=1.93, P=0.05).

  
  
• Any change in visual acuity [ Time Frame: 3 year extension, total 5 year study from the baseline ] [ Designated as safety issue: No ]
  Loss of ten or more letters was found in 6/33 (18%) initial-triamcinolone eyes compared with 8/34 (24%) initial-placebo eyes.
  
  
• Number of laser treatments required. [ Time Frame: 3 year extension study, total 5 year study from baseline ] [ Designated as safety issue: No ]
  During the third to fifth years of the study, a similar proportion of eyes from the 2 groups had macular edema that warranted laser treatment: initial-triamcinolone, 5/29 (17%); initial-placebo, 6/28 (21%).
  
  

Drug: Triamcinolone acetate
When indicated, intravitreal triamcinolone (0.1 ml of Kenacort 40© [40mg/ml triamcinolone acetonide, Bristol-Myers Squibb pharmaceuticals, Australia]) was injected into the vitreous under sterile conditions in a minor procedures area.
Other Name: Kenacort 40©

A 25 fold increase in the risk of going blind on diagnosis of diabetes is one of the most daunting threats that patients face. People using insulin are particularly challenged because they are unable accurately to draw up their dose of drug. Most cases of vision impairment in diabetes are due to macular oedema that persists or recurs after laser treatment. There are now a number of uncontrolled, anecdotal reports that intravitreal triamcinolone (IVTA) is highly effective for the treatment of diabetic macular edema which is refractory to conventional laser treatment. We commenced the first placebo-controlled, double masked clinical trial of IVTA for refractory macular oedema in 2002. The 3 month results from this study provide the first scientific proof of principle that IVTA reduces macular thickness and improves vision. The two year results will be available in March 2005, but confidential interim analysis of efficacy data in September 2004 suggested that the beneficial effect of triamcinolone treatment persisted. Thus it appears that treatment with IVTA may be the most significant development for the prevention of blindness in people with diabetes since the introduction of laser treatment. It would also be a highly cost-effective intervention that could be administered by general ophthalmologists. The treatment cannot be recommended for routine use, however, until its long term efficacy and safety have been established. Since we already have a well studied group of patients who have received treatment for 2 years, we are in a unique position to extend the study in order to provide the long-term (5-year) safety and efficacy data that does not appear to be forthcoming from any other source. The completion of this study will have a direct and immediate effect on the risk of blindness in people with diabetes by allowing doctors to predict more accurately the long term effects of this promising new treatment.