The Use of Optical Coherence Topography to Predict the Response to Argon Laser Treatment of Diabetic Maculopathy

Abstract

Purpose: : To assess if optical coherence tomography (OCT) demonstrates the effects of focal argon laser, can be used as a guide for treatment and predict response to laser earlier than the standard 3-month wait.

Methods: : Prospective, consecutive, non-intervention observational case series of patients requiring unilateral focal argon laser treatment for clinically significant macular oedema (CSMO) for the first time. Serial retinal thickness measurements pre and post laser. Outcome measures: qualitative and quantitative assessment of retinal thickness with OCT-mapping analysis by identifying OCT-oedema (OCT-O) and sectors lasered, classifying OCT-O as OCT-CSMO (using ETDRS definition of retinal thickness), predicting response to laser by qualitative change in OCT colour-coded maps, best-corrected visual acuity (BCVA).

Results: : 26 patients had serial retinal thickness measurements with OCT prelaser and every 2 weeks for 12 weeks. OCT-O was evident on 89% of prelaser maps of treated eyes. Prelaser, compared with clinical assessment, classification of OCT-CSMO had sensitivity 81% and specificity 62%. Fellow control eyes with OCT-O, were significantly different to treated eyes with OCT-O with sector retinal thickness threshold of 325 µm, p=0.001. Quantitative analysis of OCT-O sectors lasered, showed a significant decrease in retinal thickness from prelaser from 6 weeks, -15.1 µm (95%CI: -29.4 to -0.8) with a larger decrease at 12 weeks of -26.2 µm (95%CI: -38.8 to -13.5), p=0.0002. There was no significant change for the OCT-O sectors not lasered. Increase in logMAR BCVA for treated eyes was less than 0.1 (5 letters). Expert and inexperienced raters had very good intra-rater (Κ^w = 0.83, 0.78, respectively) and good inter-rater (Κ^w = 0.661) agreement for the prediction of responders to laser treatment by analyzing OCT topographic maps. Prediction analysis for responders was greatest at 8 weeks for both expert (sensitivity 88%, specificity 89%) and inexperienced rater (sensitivity 80%, specificity 88%).

Conclusions: : OCT could be used to detect patients likely to require further treatment for CSMO at 8 weeks. There was moderate correlation of thickening on OCT compared to clinical examination. Laser to the areas of thickening as seen on OCT responded best.

Keywords: diabetic retinopathy • imaging/image analysis: clinical • laser