European Association for the Study of Diabetic Eye Complications

Annual Meeting Rome 26-27th May 2007

Free Paper Abstracts 13 - 22

Saturday 26th May 2007

Free Paper Session: 14:30 - 15:45 continued

Free Paper Abstract 13

REACHING THOSE WHO NEED IT THE MOST – DIABETIC RETINOPATHY SCREENING IN EAST LONDON. T Peto,¹ A Powling,² C Schmermer.
¹ Moorfields Eye Hospital and Tower Hamlets PCT,² London, UK

Purpose: The population of patients with diabetes doubled since 1995 in Tower Hamlets (TH), an underprivileged East London Borough where the estimated prevalence is now 6.1%. In addition, TH had the highest diabetes related mortality in the UK in 2001. It was believed that all diabetes related problems were exacerbated by cultural and language barriers as TH houses the largest Bengali population outside of Bangladesh. It was estimated that attendance at diabetic retinopathy screening DRS was only 30% in 2001.
Methods: In 2001, the TH Diabetes Team, including DRS, was formed to improve patient services. Systematic recruitment of staff and patients was moved forward by a partnership between TH and Moorfields Eye Hospital, London, UK. A team, comprising of a clinician, screeners and administrative personnel was formed to improve attendance in DRS. The accuracy of the database was determined first by cross-examining relevant hospital databases. Second, an educational campaign was carried out; this aimed at physicians, diabetologists and religious leaders of the community. Third, appointments were booked and patients were screened by Bengali-speaking staff to minimise the anxiety associated with screening. Furthermore, a second clinic in a geographically more convenient location was set up to facilitate take-up of services.
Results: The DRS database was found to be over 90% accurate. On days when Bengali-speaking administrators made bookings, the attendance rate increased dramatically (from 50% to 75%) this for follow-up patients was 90%. On hospital databases a further 202 patients were identified as participating in slit-lamp screening. Bengali speaking screeners at the second location brought the overall attendance rate to 70%.
Conclusions: A team approach coupled with administrative and educational actions was effective in raising participation in the TH DRS 72% of eligible patients screened to date. This was then within government guidelines for DRS in the UK. Further improvement is needed if current government guidelines are to be met.

Free Paper Abstract 14

VISUAL ACUITY IN DIABETICS AT FIRST SCREENING FOR EYE DISEASE DIABETIC RETINOPATHY SCREENING SERVICE FOR WALES (DRSSW). H Cilliers, DR Owens, A Reidy. Royal Gwent Hospital, Newport, Wales

Introduction: Diabetic retinopathy is the third major cause for visual impairment in the UK and screening services attempt to reduce the incidence of severe blindness by early detection in this high risk population group.
Aims: The aims of this study were to measure visual acuity impairment in a cohort of diabetics at first presentation for screening at the Diabetic Retinopathy Screening Service for Wales (DRSSW) and to establish the causes thereof. A further aim was to estimate the numbers of visually impaired diabetics whose visual loss is remediable or preventable, to be able to project the impact of that workload on Ophthalmic Services.
Methods: Data was used from diabetics attending for their first screening at the DRSSW. After their photographs had been graded and data collated, an estimate of the prevalence and distribution of visual impairment was performed according to age and disease. Multiple regression analysis of risk factors for disease was done by statistical analysis to evaluate risk factors for visual impairment.
Results: Of the 27,178 diabetics, 4.4% (1,193) had moderate visual impairment and 1.4% (393) had severe visual impairment. Multiple logistic regression analysis of risk factors for visual impairment were female sex – Odds ratio (OR) 1.6, [95% confidence interval (CI) 1.4-1.8]; aged 65 years or older – OR 2.7 [95% CI 2.4-3.0]; proliferative diabetic retinopathy – OR 5.4, [95% CI 3.7-7.7] and diabetic maculopathy – OR 1.7 [95% CI 1.4-2.0]. The proportion of individuals who required eye care services for visual impairment was estimated at 6.3% (1,703).
Conclusion: With this analysis the DRSSW can provide projected numbers with regards to visual impairment in the diabetic population and the resulted impact of these on eye care services. With the incidence of diabetes rising, diabetic eye screening to optimise Health Care provision can aid future planning and cost structuring.

Free Paper Abstract 15

DIABETIC RETINOPATHY SCREENING WITH FUNDUS PHOTOGRAPHS: COMPARISON OF DIFFERENT APPROACHES. E Pilotto, S Vujosevic, E Benetti, F Massignan, E Midena. Department of Ophthalmology, University of Padova, Italy

Purpose: To analyse the ability to identify individual lesions and to determine clinical levels of diabetic retinopathy using non-mydriatic digital colour retinal images compared to Early Treatment Diabetic Retinopathy Study (ETDRS) seven standard 35-mm stereoscopic colour fundus photographs (ETDRS photos).
Methods: Sixty-six eyes of diabetic patients with type 1 or type 2 diabetes mellitus from a single center including the full spectrum of diabetic retinopathy were enrolled. Non-simultaneous 45°-field non stereoscopic digital colour images were taken from one central field and from three fields with a non-mydriatic fundus camera (Nidek Technologies, Albignasego, Italy) (1F-NM and 3F-NM images) before pupil dilation. The 3F-NM procedure for field location was automatically performed by the fundus camera. Following pupil dilation, standard 30° ETDRS photos were obtained with Topcon TRC50IA (Tokyo, Japan). NM images and ETDRS photos were graded on a lesion-by-lesion basis by two independent, masked readers to assess the ETDRS grading of the lesions, and the clinical severity of the disease according the International proposed five-stage disease severity Classification.
Results: Comparison of individual lesions among 1F-NM image or 3F-NM images and ETDRS photos revealed good agreement for hard exudates (k=0.74 and k=0.75 respectively); moderate agreement for haemorrhages and/or microaneurysms (k=0.49, k=0.60), cotton wool spots (k=0.41, K=0.49) and retinal new vessels (k=0.40, k=0.43). The agreement was fair for clinical significant macular edema (k=0.40, k=0.39). There was very high agreement between 3F-NM and ETDRS images for severe non-proliferative and proliferative diabetic retinopathy (k=0.81 and k= 0.82 respectively), while it was moderately good between 1F-NM and ETDRS images.
Conclusions: Undilated digital images using non-mydriatic cameras are becoming a new tool for the screening of diabetic retinopathy. The agreement appeared excellent for high risk diabetic retinopathy levels, while it was poor for the clinical significant macular oedema due to non-stereo technique. The automatic non-mydriatic system may be an effective tool for determining the level of the diabetic retinopathy and identifying the need for prompt referral to the ophthalmologist, even when applied as a telemedicine tool. It is still controversial if one 45° image instead of three fields images is adequate for diabetic retinopathy screening.

Free Paper Abstract 16

PROLIFERATIVE DIABETIC RETINOPATHY IS AN IMPORTANT PREDICTOR OF MORTALITY AMONG TYPE 1 DIABETIC PATIENTS. J Grauslund,¹ A Green,² AK Sjølie.¹
Dept. of Ophthalmology ¹ and Dept. of Applied Research and Health Technology Assessment,² Odense University Hospital, Odense, Denmark.

Purpose: To evaluate the effect of diabetic retinopathy on 25-year survival among a population-based cohort of type 1 diabetic patients from the Danish County of Fyn.
Methods: In 1973 all type 1 diabetic patients from Fyn County, Denmark, with onset before the age of 30 as of July 1st 1973 were identified (n=727). In 1981-1982 573 of the 627 patients still alive and living in Denmark participated in a clinical examination where diabetic retinopathy was graded and other markers of diabetes measured. 25-years follow-up mortality was examined in 2006 and related to the baseline examination.
Results: Of the 573 patients participating in the baseline examination in 1981-1982 297 (51.8%) were still alive. 256 had died (44.7%). 3 had emigrated (0.5%) and 17 (3.0%) were unaccounted for because they had chosen not to provide data for scientific examinations. There was no difference in survival among males and females (p=0.52). Age- and sex-adjusted hazard ratios (HRs) of mortality were 1.01 (95 CI: 0.72-1.42) and 2.04 (1.43-2.91) for patients with non-proliferative and proliferative retinopathy at baseline compared to patients with no retinopathy, respectively. In a univariate model different risk factors measured at baseline were added to the above model. Adjusting for diabetes duration, smoking, HbA1c, systolic and diastolic blood pressure and BMI only changed HR marginally. After adjusting for proteinuria however, HR decreased to 1.49 (0.99-2.24). In a multivariate model including all above risk factors, HR among patients with proliferative retinopathy was no longer statistically significant but still remaining increased 1.48 (0.98-2.23).
Conclusions: Proliferative diabetic retinopathy is a predictor of mortality among type 1 diabetic patients. This association is partly explained by proteinuria but HR still remains increased after adjusting for proteinuria. Nonproliferative diabetic retinopathy at baseline did not affect survival among the type 1 diabetic patients in the current study and there was no difference in mortality between men and women.

Sunday 27th May 2007

Free Paper Session: 09.00 – 10.30

Free Paper Abstract 17

DIABETIC MACULAR OEDEMA: EFFECTS OF CHANGES IN PLASMA OSMOLARITY ON THE RETINAL THICKNESS AS EVALUATED BY OCT. DN Thornit, B Sander, H Lund-Andersen. Department of Ophthalmology, Glostrup Hospital, University of Copenhagen, Denmark.

Purpose: To examine if the osmotic Starling forces affect diabetic macular oedema by analysis of the effect of increasing the plasma osmolarity on the retinal thickness in clinically significant diabetic macular oedema (CSMO).
Methods: Prospective, paired, unmasked study. On two separate days, 14 patients with diabetic CSMO were randomised to an oral glycerol solution (0.57 g/mL) of 1.5 and 3 mL/kg body weight (maximum 250 mL). Plasma osmolality (p-OSM) and glycerol (p-GLY); and the primary effect parameter: the retinal thickness (RT) of the peak oedematous field on the retinal maps of the StratusOCT™ fast protocol, were monitored for the subsequent 180 min. The systemic blood pressure (BP), intraocular pressure (IOP) and capillary glucose (CG) were also monitored over the same time interval.
Results: P-OSM and p-GLY correlated strongly (r=0.79, p<0.0001). Baseline p-OSM was 300 (4) mOsm/L (mean)(SD) for the low and 302 (8) mOsm/L for the high glycerol dose. The peak &#916; p-OSM increased from 13 (9;17) mOsm/L (mean)(95% confidence limits, CL) at 60 min to 23 mOsm/L (20;27) mOsm/L at 120 min, respectively (p<0.0001). Baseline RT was 377 (73) (mean) (SD) and 373 (68) µm for the low and high dose, respectively. The maximal &#916; RT of 7 µm (4;10 µm) (mean) (95% CL) (p=0.0002) after the low dose and 7 µm (3;12 µm) (p=0.006) after the high dose occurred in the interval from 20 to 30 min, where the dose effect of p-OSM had not yet become significant.
Neither BP, IOP nor CG interacted significantly with the thickness results (p>0.05).
Conclusions: Glycerol transiently reduces the peak RT in diabetic CSMO confirming the application of the osmotic Starling forces on diabetic macular oedema; however, no dose dependent effect was found.

Free Paper Abstract 18

ASSESSMENT OF DIABETIC MACULAR EDEMA: A COMPARISON OF DIFFERENT METHODS. NN Grigoryeva, EB Shklyarov, YS Astakhov, FE Shadrichev. St-Petersburg State Pavlov Medical University, Russia

Purpose: To compare effectiveness of 35° stereo fundus photography, optical coherence tomography (OCT), confocal retinal tomography (Heidelberg retina tomography II – HRT II), fluorescein angiography (FA) with non-contact lens biomicroscopy for the diagnosis of diabetic macular edema (DME).
Methods: 136 patients (210 eyes) with DME, 30 diabetic patients without macular edema (54 eyes), 35 healthy subjects (67 eyes) were examined using with non-contact lens biomicroscopy, stereophotography (Topcon TRC-50IX fundus camera, Agfachrome 100 colour film), OCT (Stratus 3000, scan protocol «Macular Thickness»), HRT II (Macula Edema Module, version 1.6), FAG (Topcon TRC-50IX fundus camera, IMAGEnet 2000).
Results: In comparison to fundus biomicroscopy the following sensitivity and specificity were found: for OCT – 0.98 and 0.85, k – 0.82; for HRT II – 0.94 and 0.76, k – 0.71; for FAG – 0.90 and 0.92, k – 0.76 and for stereophotography – 0.85 and 0.89, k – 0.75 (p < 0.001) respectively. OCT and FAG were the optimal combination (OCT and FAG – sensitivity 0.98 and specificity 0.87, k – 0.84, p < 0.001; OCT and HRT II – 0.98 and 0.70, k – 0.69, p < 0.001; HRT II and FAG – 0.60 and 0.55, k – 0.60, p < 0.001, respectively). The agreement between methods was fair (OCT and FAG – k 0.56; OCT and stereophotography k - 0.59; FAG and stereophotography k – 0.46).
Conclusions:The best agreement with fundus biomicroscopy was shown by OCT. Combination of OCT and FAG is more informative for DME diagnosis, because it allows the assessment of morphological and physiological criteria of DME.

Free Paper Abstract 19

MULTIFOCAL ERG RESPONSE SEEMS TO IMPROVE IN AREAS TREATED WITH PHOTOCOAGULATION FOR DIABETIC MACULAR OEDEMA. K Holm, M Lovestam-Adrian. Department of Ophthalmology, University Hospital, Lund, Sweden.

Purpose: To evaluate the local correspondence between multifocal electroretinogram (mfERG) response and retinal thickness assessed with OCT after focal laser treatment in patients with diabetic macular oedema.
Methods: Twelve diabetic patients (aged 60±14 years, diabetes duration 16 ±8 years) treated with focal/grid photocoagulation for clinically significant macular oedema underwent mfERG and optical coherent tomography (OCT) before and three months after treatment. The fixation during mfERG recording was controlled using a fundus camera and illumination with infrared light from the recording electrode, with visualization of the hexagonal elements over the retina. The average thickness (µm) in any of the nine sectors, as defined by the ETDRS, which was treated with photocoagulation was measured. Amplitudes and implicit times were analyzed within corresponding areas on the mfERG.
Results: There was a borderline increase in mfERG amplitudes after photocoagulation; 20.4 ±7.5 vs. 15.8 ±6.2 nV/deg2; p=0.055, whereas no difference was seen in implicit time. OCT values in the treated regions were lower at follow-up 272 ±23µ vs. 327 ±79µ; p=0.013. No correlation was seen between the changes in mfERG response and changes in OCT values. The decrease in retinal thickness was correlated with the amount of laser spots given p=0.002.
Conclusion: Focal argon laser treatment is effective in reducing retinal thickness and there is a tendency toward improved retinal function in treated areas demonstrated by increased amplitudes on the mfERG.

Free Paper Abstract 20

OCT FEATURES DURING EVOLUTION OF SEROUS RETINAL DETACHMENT ASSOCIATED WITH DIABETIC MACULAR EDEMA. D Gaucher, C Sebah, A Erginay, B Haouchine, R Tadayoni, A Gaudric, P Massin. Hopital Lariboisière, Paris, France

Purpose: To analyse the evolution of serous retinal detachment (SRD) associated with diabetic macular edema (DME) using optical coherence tomography (OCT).
Methods: 64 eyes of 40 diabetic patients who had SRD associated with DME were studied retrospectively. All the patients had fluorescein angiography and repeated OCT3 examinations during follow up. Foveolar neuroretinal thickness (NRT) and height of the SRD (HSD) were measured. The evolution of the OCT macular profiles was qualitatively analysed.
Results: Mean follow-up was 11.8 months. DME was focal in 10 eyes (15.6%), diffuse in 17 (26.6%) and both diffuse and focal in 37 (57.8%). Mean (+/-SD) initial VA, NRT and HSD were 0.35 (+/- 0.21), 346.88+/- 138.61 µm and 199.48+/-139.8 µm respectively . HSD was not correlated with VA (p=0.23) nor with NRT (p=0.31). In 9 eyes (14.1%), NRT above the SRD was normal. In eyes in which DME improved during follow up(19 eyes), SRD disappeared before the maximal reduction of retinal thickness in 7 eyes (36.8%) and after or simultaneously in 12 (63.2%). In eyes in which DME worsened during follow up (45 eyes), the SRD disappeared in 15 (33.3%).
Conclusion: In this study, SRD height did not correlate with retinal thickening. The latter may appear before central neuroretinal thickening and disappear before or after its regression. Consequently, SRD does not seem to be related either to the severity of the DME nor to its reabsorption.

Free Paper Abstract 21

INTRAVITREAL TRIAMCINOLONE VERSES LASER PHOTOCOAGULATION FOR DIABETIC MACULAR OEDEMA. Z Ockrim, S Sivaprasad, S Falk, C Bunce, Z Gregor, P Hykin. Moorfields Eye Hospital, London, UK

Purpose: To determine if repeated intravitreal triamcinolone improves best corrected visual acuity at 1 year compared to conventional laser therapy for persistent diabetic macular oedema.
Methods: This was a prospective randomised controlled clinical trial. 88 eyes with persistent clinically significant macular oedema after at least one prior laser photocoagulation. 43 patients were randomised to intravitreal triamcinolone and 45 to laser photocoagulation 4mg of 0.1ml intravitreal triamcinolone versus standardised ETDRS macular laser photocoagulation were given every 4 months for one year. The primary endpoint was the proportion of patients who improved by 15 ETDRS letters at 12 months. Secondary endpoints were the change in mean best corrected visual acuity at 12 months, difference in macular thickness and macular volume in TA vs laser groups and adverse event reporting in particular elevated intraocular pressure.
Results: Improvement in 15 or more ETDRS letters was seen in 2 out of 41 patients in the ivTA group (4.9%) and in 5 out of 41 (12.2%) patients in the laser group (p = 0.492). The mean Early Treatment of Diabetic Retinopathy Study (ETDRS) scores at baseline were 54.4 letters in the study group and 53.0 letters in the control group (p = 0.59). At 12 months the mean ETDRS scores were 54.5 and 54.6 respectively (p = 0.97). Optical coherence tomography showed a reduction in central macular thickness with triamcinolone of 82.0 um and 62.3 um with laser (p = 0.6) at 12 months. There was 1 case of sterile endophthalmitis. 22 out of 43 patients in the study group required ocular antihypertensives.
Conclusion: This study did not show any benefit from intravitreal triamcinolone for patients with persistent diabetic macular oedema compared to conventional laser therapy and its use is not recommended in routine clinical practice.

Free Paper Abstract 22

INTRAVITREAL BEVACIZUMAB (AVASTIN) FOR DIABETIC MACULAR EDEMA. DR Pognuz, F Menchini, F Bandello. Department of Ophthalmology, University of Udine, Italy.

Purpose: To report the safety and efficacy of intravitreal injections of bevacizumab for the treatment of diabetic macular edema.
Methods: This prospective, non-comparative case series included 21 eyes of 21 consecutive patients with diabetic macular edema. Patients with macular edema, independently of the size and duration of edema, retinal thickness, visual acuity and type of diabetes were included. Among the exclusion criteria were any prior treatments (either macular laser or intravitreal triamcinolone) in the study eye within 6 months preceding enrolment, presence of non-perfusion for >1 disc area involving the foveal avascular zone, epiretinal membrane and significant media opacities. At each visit all patients underwent assessment of best corrected visual acuity (BCVA) with ETDRS-like charts, slit-lamp examination, intraocular pressure measurement, retinal thickness measurement using optical coherence tomography (OCT), fluorescein angiography and fundus biomicroscopy. Main outcome measures were occurrence of treatment-related ocular or systemic complications, mean change in BCVA, change in foveal thickness on OCT. All patients received at least one intravitreal injection of 1mg/0.04 ml bevacizumab (Avastin) on a monthly basis.
Results: All patients completed 3 months of follow-up. 16 patients (76%) received 3 injections, 3 patients (14%) and 2 patients (10%) received 2 and 1 injection respectively. All patients but two had undergone previous treatments, such as grid laser, full scatter panretinal photocoagulation and intravitreal injection of triamcinolone with no benefit or with recurrence of the edema. A total of 134 injections were performed with no evidence of systemic or ocular side effects. At baseline, mean BCVA was 59.4 ±14.2 ETDRS letters, ranging from 34 to 84, and mean central retinal thickness by OCT was 501 ±146 µm (range 321-721 µm). Mean BCVA at 4 months was 64.5 ±8.5 ETDRS letters: changes in visual acuity were not significant throughout follow-up. Mean retinal thickness decreased to 444 ±177 µm, 456 ±59 µm and 426 ±192 at 4, 8 and 12 weeks respectively. Anatomical improvement was sustained up to the 4th month and was statistically significant.
Conclusions: Preliminary, short-term results suggest that IVB provides morphological improvement with no significant adverse events in patients with diabetic macular edema. Decrease in mean retinal thickness was significant but it was not associated with an overall increase in mean BCVA, except for 3 patients with severe non proliferative diabetic retinopathy and moderate to marked peripheral retinal ischemia. This seems to suggest that in such patients VEGF may play a predominant role, compared to other mediators, in the genesis of the edema and its inhibition could provide both functional and anatomical improvement.