Aim: To determine if repeated intravitreal triamcinolone improves best corrected visual acuity at 1 year compared with conventional laser therapy for persistent diabetic macular oedema.
Methods: 88 eyes with persistent clinically significant macular oedema, after at least one prior laser photocoagulation, were included in this prospective randomised controlled trial. 43 patients were randomised to 4 mg of intravitreal triamcinolone (TA) and 45 to laser photocoagulation. The primary endpoint was the proportion of patients who improved by 15 Early Treatment of Diabetic Retinopathy Study (ETDRS) letters at 12 months in TA versus laser groups. Secondary endpoints were the change in mean best corrected visual acuity, difference in macular thickness and macular volume and adverse event reporting in particular elevated intraocular pressure at 12 months.
Results: Improvement in ⩾15 ETDRS letters occurred in two of 42 patients in the TA group (4.8%) and in five of 41 (12.2%) patients in the laser group (p = 0.265). At baseline, the mean ETDRS scores at baseline were 54.4 letters in the TA group and 53.0 letters in the laser group. At 12 months, these were 54.5 and 54.6, respectively. Optical coherence tomography showed a reduction in central macular thickness of 82.0 μm with TA and 62.3 μm with laser at 12 months. There was one case of sterile endophthalmitis. 22 out of 43 patients in the triamcinolone group required ocular antihypertensives.
Conclusion: This study did not show a benefit from intravitreal triamcinolone over conventional laser therapy for patients with chronic diabetic macular oedema
Statistics from Altmetric.com
Diabetes mellitus is the commonest cause of blindness in the working population in England and Wales, with over 80% due to diabetic maculopathy.1 2 The prevalence of diabetic maculopathy is directly related to the duration of diabetes, with approximately 30% of patients with diabetes affected after 25–30 years of disease. As the incidence of diabetes increases, the rate of diabetic complications including diabetic macular oedema and visual loss is set to increase.3
The only proven effective treatment for diabetic macular oedema is laser photocoagulation.4 However, this treatment only reduces the chance of moderate visual loss by 50% and is unlikely to improve visual acuity. As a result, better treatments have been sought.
Initial pilot studies using intravitreal triamcinolone (TA) suggest it can improve visual acuity in patients with diabetic macular oedema in approximately 30% in the shorterm.5 6 However, long-term evidence of efficacy from prospective data is limited. The aim of this study was to determine if repeated TA improves the Early Treatment of Diabetic Retinopathy Study (ETDRS) best corrected visual acuity (BCVA) letter score at 1 year compared with conventional laser therapy and whether this treatment has a role to play in the long-term care of patients with diabetic macular oedema.
PATIENTS AND METHODS
Patients were recruited via the Medical Retinal Service at Moorfields Eye Hospital from February 2003 to June 2005. If both eyes were eligible for inclusion, the eye with the worst BCVA was included, and the fellow eye followed standard treatment.
Inclusion and exclusion criteria are summarised in table 1. Clinically significant macular oedema (CSMO) as defined by the ETDRS4 was diagnosed on slit-lamp biomicroscopy and confirmed angiographically and tomographically.
Sample size calculation
The principal outcome measure was the proportion of patients who improved by 15 ETDRS letters at 12 months. Based on the original pilot studies5 6 the estimated frequency in the TA arm would be approximately 20% and from the ETDRS in the laser arm approximately 3%. The study was powered at 80%, with a significance level of 0.05. Therefore, for a ratio of study patients to control patients of 1:1, the calculated sample size was 41 patients in each arm. Adjusting for an estimated loss to follow-up of 8% of eyes, we aimed to recruit 44 patients into each group.
BCVA was measured by a trained optometrist, masked to treatment arm, using an ETDRS chart to record distance vision, a Bailey–Lovie chart at 25 cm to record near vision and a Pelli–Robson chart at 1 m to record contrast sensitivity. Intraocular pressure (IOP) was measured using Goldman applanation tonometry. Funduscopy was performed by slit-lamp biomicroscopy using a 66D lens. The degree of lens opacity was graded using the Wisconsin system for classifying cataracts.7
The maximal linear diameter of the foveal avascular zone (FAZ) was calculated using the measuring tool on the Topcon imaging system and amount of capillary dropout assessed against the ETDRS standard photograph.
Optical coherence tomography
OCT was performed by a certified research technician, masked to the treatment arm. The macula was scanned using the 6×6 mm radial line scan protocol using the Stratus OCT model 3000 OCT scanner (Zeiss Humphrey Instruments, Dublin, CA). All computer-generated measurements were visually inspected by the study investigator to exclude computerised artefacts. Foveal minimum was selected to represent the central macular thickness from the computer-generated retinal thickness analysis program, and macular volume, where possible, was calculated using the computer algorithm.
Randomisation and treatment
Patients were allocated to 4 mg of TA or further ETDRS laser photocoagulation using a centralised computer-generated randomisation program (based on the weighted coin method) in operation at the Clinical Trials Unit. Patients were retreated at 4 and 8 months if they had persistent macular oedema on clinical examination with final review at 12 months.
Baseline treatments took place within 1 week of the screening visit. Intravitreal triamcinolone was performed by the study investigator in a minor procedure area in the outpatient clinic under sterile conditions. The eyes were prepared with topical drops of 1% amethocaine and a 5% povidone-iodine flush. A subconjunctival injection of 2% lignocaine was given before 4 mg (0.1 ml) of triamcinolone acetonide (Kenalog, Bristol Myers Squibb, Hounslow, UK) was injected through the pars plana into the mid-vitreous without removing the vehicle. Following injection, indirect ophthalmoscopy was used to check central retinal artery perfusion, and if occluded, a paracentesis was performed. Topical chloramphenicol 0.5% was instilled. Following the procedure, eyes were checked at the slit lamp, to look for triamcinolone crystals in the anterior chamber, to examine the fundus and to measure IOP. Patients were discharged home with topical chloramphenicol 0.5% qds for 4 days. Postoperative checks were repeated at 1 and 4 weeks.
Conventional macular laser was performed by the study investigator based on ETDRS guidelines.4 In summary, a pretreatment fluorescein angiogram was used to identify areas of leakage. Areas were treated with 100 µm argon green-only burns of 0.1 s duration, with adequate power to obtain definite whitening in areas of leakage. Care was taken to avoid the FAZ.
The primary outcome measure was the proportion of patients who improved by 15 or more ETDRS letters at 12 months. Secondary endpoints were to determine if TA improved mean ETDRS letter score at 12 months reduced mean central retinal thickness and macular volume compared with conventional laser therapy and adverse event reporting. Adverse event reporting included a change in IOP measurements and a significant change in the size of the FAZ.
We analysed the data according to the intent to treat principle. However, data for five (5.7%) of the patients (four of whom were on laser treatment) were not available on the primary outcome measure at 1 year. We did not consider it valid to carry forward data from earlier visits and thus adopted the available case analysis approach. Baseline characteristics of the patients were studied in order to assess the adequacy of randomisation. The Fisher exact test was used to assess the statistical significance of the difference in the proportion of patients in each group who improved by 15 or more ETDRS letters. Analysis of covariance was used to assess secondary outcome variables.
A total of 88 eyes from 88 patients were included, of which 43 eyes were randomised to receive intravitreal triamcinlone and 45 eyes to receive laser. Data for 1 year were available in 83 patients (94%). Five patients did not complete follow-up. Two patients died, one patient moved away from the area, one patient was unable to travel to City Road, and one patient (who received triamcinolone) refused follow-up.
The baseline characteristics of the patients in each group are shown in tables 2 and 3, which show that the groups were comparable with regards to all characteristics assessed.
Eighty-one injections were performed over the 12 months (mean = 1.9 treatments per patient). In the same period, 73 laser treatments were performed (mean = 1.6 treatments per patient).
An outline of the injection profile for the patients undergoing intravitreal TA is summarised in table 4. Ten of 43 patients received all three injections. Twenty-two injections (17%) were not performed due to a resolution of the CSMO on clinical examination. Six injections (5%) were not performed either when the patient failed to attend the appointment (four treatments) or when treatment was refused (two treatments). Eighteen injections (14%) were not performed due to protocol exclusions which included: (1) a rise in the IOP (13 treatments (10%)); (2) dense cataracts obscuring the fundal view (two patients); (3) one patient (two treatments) was deemed unfit for further injections following a myocardial infarction just prior to the second injection; and (4) one patient had a sterile endophthalmitis following the second injection.
A change in the visual acuity letter score at 12 months is presented in figure 2. Improvement in 15 or more ETDRS letters was seen in two of 42 patients in the TA group (4.8%) and in five of 41 (12.2%) patients in the laser group (p = 0.265).
At baseline, the mean ETDRS letter scores were 54.6 in the triamcinolone group and 53.0 in the laser group. At 12 months, they were 54.4 and 54.7, respectively, and analysis of covariance indicated no evidence of a treatment effect (p = 0.44) (table 5)
Similarly, we found little evidence of any differences in the mean near acuity (Bailey–Lovie scores) or contrast sensitivity (Pelli–Robson scores)
Retinal thickness and macular volume
The changes in central retinal thickness (RT) and macular volume are presented in figures 3 and 4. At baseline, the mean RT was 413.4 μm in the study group and 410.4 μm in the control group. At 4 months, the mean RT had fallen to 325.6 μm in the study group and 374.2 μm in the control group. At 12 months, the mean RT was 322.1 μm in the study group, and in the laser group the mean RT had fallen to 346.7 μm. Even the largest difference observed (ie, that at 4 months) was not statistically significant in these data (p = 0.19), and ANCOVA revealed no evidence of a difference (0.738)
At baseline, the mean macular volume was 9.74 mm3 in the triamcinolone group and 9.42 mm3 in the laser group. At 4 months, the mean macular volume had fallen to 8.45 mm3 in the triamcinolone group and 9.08 mm3 in the laser group. At 12 months, the mean macular volume was 8.41 mm3 in the triamcinolone group, and the laser group had reduced to 8.66 mm3. Again, ANCOVA revealed no evidence of a difference (p = 0.65)
There was one case of sterile endophthalmitis (one case in 78 injections (1.3% of injections, 2.3% of patients)). This patient was reviewed at 1 week following his third intravitreal injection with a drop in VA. The eye was white, with +1 of cells in the anterior chamber, and a hazy vitreous. A vitreous biopsy was negative, and despite intravitreal antibiotics there was little change in the appearance of the vitreous 12 months after antibiotics.
There were three cases in two eyes of triamcinolone in the anterior chamber. All cases were detected immediately following injection and settled spontaneously within 2 weeks of injection. The IOP remained within normal limits.
The most common adverse event was a rise in IOP (table 6). There was a rise in the mean IOP in the study group from 15.9 mm Hg to 20.3 mm Hg at 4 months. Furthermore, 22 out of 43 patients required ocular antihypertensives at some point during follow-up. However, all rises in IOP were easily managed with one or more ocular antihypertensives, and further interventions such as vitrectomy to remove the TA or glaucoma drainage procedures were unnecessary. There were no cases of glaucoma following TA.
The change in the degree of lens opacity for nuclear sclerosis, cortical lens opacity and posterior subcapsular lens opacity (PSCLO) is presented in fig 5. The only statistically significant change was in the degree of PSCLO in the triamcinolone group compared with the laser group. The degree of lens PSCLO changed in the triamcinolone eyes from a mean of 1.1 to 1.9, while the laser eyes changed for a mean of 1.1 to 1.3.
Two eyes in the triamcinolone group and one eye in the laser group underwent additional cataract surgery for visually significant lens opacity.
Finally, there was no significant change in the mean size of the foveal avascalar zone in either group from baseline to 12 months (study = 758 to 759 μm, control = 682 to 762 μm).
To our knowledge, there has been no previous prospective randomised trial that directly compares TA with argon laser photocoagulation for patients with persistent CSMO, and so this study was aimed to address this. Triamcinolone is unlikely to gain acceptance in clinical practice unless it confers a significant advantage over conventional laser photocoagulation, and consequently the primary endpoint was the proportion of patient in each group with improved visual acuity. The sample size calculations were based on a 17% difference which was calculated from the original two pilot studies.
This study did not achieve the primary endpoint of an improvement of 15 ETDRS letters at 12 months in the triamcinolone compared with the laser group. Preliminary studies showed an improvement in VA and a reduction in central retinal thickness in CSMO. However, these studies were typically uncontrolled. There was a small reduction in central macular thickness in the short-term following triamcinolone in this study, although this was not statistically significant from the laser group. It is possible that our study was underpowered to show a difference. This seems unlikely, as no difference was noted from the controls at any time point, and indeed at times there seemed to be a better result from the controls than from the TA group.
Gillies and co-workers recently published a beneficial effect of triamcinolone in visual acuity at 2 years.8 This was a prospective randomised study in which the primary endpoint was a five-letter gain compared with the 15-letter gain in this study. Furthermore, 50% of their patients had cataract surgery in the second year of study, and these two factors may explain, at least in part, the different study outcomes. Interestingly, this study showed progression of posterior subcapsular cataract in the triamcinolone group, and although the change was small it was statistically significant.
Similarly, Avitabile and colleagues found a positive effect on both visual acuity and central retinal thickness in eyes treated with TA and TA combined with laser.9 Despite the effect of the TA regressing at 6 months and some patients requiring a second injection, the results of TA were in marked contrast with laser. Eyes treated with laser alone had no change in either VA or central retinal thickness.
The most significant difference between this study and our own is the duration of macular oedema. Our aim was to include patients with reversible macular oedema, but in reality our study population included patients with a long duration of macular oedema (median duration 27 months, IQR 15–48 months). In comparison, the duration of macular oedema was 6–12 months (mean 9 months) for the study by Avitabile and colleagues.9 Perhaps if we had excluded patients with chronic macular oedema, we would have found a significant improvement in visual acuity, although the population of patients selected for this trial is likely representative of patients who undergo intravitreal TA in clinical practice.
The study was designed to give repeated intravitreal injections to determine whether prolonged intravitreal triamcinolone improves visual acuity long term compared with a single injection which may have only a transient effect. It is notable that 46/129 injections were not performed (28 of the 46 were the third injection). It could be postulated that the significant number of third injections not performed may have an adverse effect on outcome if, for instance, the first two injections had led to a significant improvement in visual acuity and reduction in central macular thickness. However, this is unlikely, as at no study time-point was there a significant improvement in visual acuity or a decrease in central macular thickness or macular volume. It may be possible to have given final injections by raising the IOP threshold to exclude fewer patients. However, by choosing a threshold of 30 mm Hg, there were no serious cases of persistent raised IOP requiring filtration surgery.
This study is relatively small, with limited follow-up. It directly compares to conventional ETDRS-recommended macular laser therapy, the only recognised standard treatment for clinically significant macular oedema. However, the patient population tended to have chronic CSMO. There were few side effects with only one case of sterile endophthalmitis and no cases of retinal detachment or glaucoma. However, neither TA nor laser treatment leads to an improvement in visual acuity or a reduction in central retinal thickness at any time point. Therefore, we did not find evidence to support the use of triamcinolone as a routine treatment for chronic CSMO. Conversely, other studies suggest a role for intravitreal triamcinolone early in the treatment for CSMO. Therefore, this study and other similar studies suggest that the treatment for CSMO should be aggressive as early as possible in order to avoid permanent structural changes.
Thank you to the Special Trustees of Moorfields Eye Hospital for funding this trial.
Competing interests: None.
Ethics approval: This prospective randomised trial was conducted in accordance with the ethical standards of the Declaration of Helsinki and was approved by the local Research and Ethics Committees at Moorfields Eye Hospital.
Patient consent: Written informed consent was obtained from all patients before details of medical and ophthalmic examinations were recorded.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.