You are here: » Vision Research » Visionary of the Quarter » Sobha Sivaprasad (Q03-2018)

We are the slaves and not the masters of OCT

Prof. Sobha Sivaprasad

The research work of Prof.Sobha Sivaprasad

In the last decade, the evolution of 3D optical sectioning of the retina using optical coherence tomography (OCT) have revolutionised our diagnostic potential of macular diseases. For the first time, retinal imaging became inevitable in the management of retinal diseases. All clinical trials in retinal vascular diseases and age related macular degeneration have since included central retinal thickness measurements in defining patient eligibility, re-treatment criteria and in stopping rules.

Despite giving importance to change in central macular thickness as an anatomical parameter in clinical trials, in routine practice, we decided to be masters of OCT and treated patients based on the morphological characteristics of the macula on OCT. So we paid significant attention to the location of hyporeflective spaces within or below the retina and defined them as intra and subretinal fluid and decided treatment based on these parameters. We even progressed to use artificial intelligence to identify these morphological features.

Legend: A Macula scan showing both intraretinal and subretinal fluid in a person with visual acuity of 73 ETDRS Letters (Snellen 20/40).

However, over time, we have realised that these spaces may in fact mean very little in terms of their correlation to visual function. With trial and error of several re-treatment criteria based on morphology of the macula on OCT, we have now started to realise that absolute retinal thickness matters more than morphology. This may be a crucial reason why real-life outcomes do not reflect clinical trial results, although we would like to believe that the sub-optimal results in real-life is due to the heterogeneity of the patient pool. 

We have also witnessed the change in treatment regimen with anti-vascular endothelial growth factor for age related macular degeneration again due to our belief that we knew better than clinical trial evidence. We started off the anti-VEGF era with pro-re-nata treatment of intra or subretinal fluid on OCT as we believed that this was the best pragmatic approach compared to fixed dosing. The OCT appearances fooled us and our outcomes were suboptimal.  We then decided fixed regimen is better and decided to ignore the OCT changes and our results closely matched clinical trial evidence.

Yet again, pragmatism overtook our judgement and we have decided to go by treat and extend regiment and base our decision on the morphology of the retina. In short, our aim is to extend when macula is dry and shorten when wet. Although we are getting better results than pro-re-nata antiVEGF treatment, the obvious question is should we be guided by fluid on OCT yet again? The integrity of the outer retinal layers is in fact far more predictive of visual outcome than any other morphological characteristics on OCT. However, overlying oedema at the time of initiation of treatment often masks these features.  

So the question is how should we be monitoring our patients in clinic? Visual acuity measurements are often inaccurate in clinic. The OCT is still the most accurate tool we have to monitor disease progression. Although OCT angiography has spiralled to be an exciting novel tool for retinal and choroidal diseases, it is no way as sensitive as OCT in assessing activity of a lesion.

The question is how could we use OCT better? Our reluctance to use change in central macular thickness compared to the last visit or the lowest value recorded previously or to baseline thickness is two fold. Firstly, we believe that we, as health professionals, should be doing better than subtracting numerical values of OCTs on serial scans. Secondly, we cannot accept the fact that morphological features in terms of presence or absence of fluid are deceptive and should not be used to determine re-treatment.

In conclusion, the OCT has stood the test of time and taught us several lessons as our master. We should learn to be the slave of the OCT and we will do much better doing our calculations on central macular thickness change than define re-treatment based on morphological characteristics on OCT.

Professor Sobha Sivaprasad


Moorfields Eye Hospital and UCL

United Kingdom

E-mail: Sobha.sivaprasad[at]