Optical coherence tomography biomarkers in preterm infants with and without retinopathy of prematurity: fallacies and facts

Dear Editor,

The recent publication from Belenje et al. [1] evaluating the optical coherence tomography (OCT) biomarkers to differentiate favourable and unfavourable responders to anti-vascular endothelial growth factor (VEGF) in retinopathy of prematurity (ROP) presented the following findings: hyperreflectivity of the inner retinal layers and choroidal thinning as indicators of unfavourable outcomes and cystic changes at the fovea as a favourable marker to anti-VEGF.

In our previous publications, we have reported using handheld OCT, the presence of inner retinal immaturity in preterm infants undergoing ROP screening. The inner retinal hyperreflectivity that Belenje et al. report here has been visualized on OCT in preterm and term infants due to the lack of or attenuation of the ganglion cell layer, a finding that has been reported in infants even without ROP [2]. (Fig. 1) Furthermore, OCT imaging technique and alignment, typically challenging to achieve with an infant in the “flying baby” position, may contribute to a ‘pseudo-hyperreflectivity’, especially in the inner retinal layers (Fig. 2).

Fig. 1. Pseudo-hyperreflectivity in an infant without ROP.

A Spectral-domain optical coherence tomography image obtained from a preterm infant with an immature retina and no ROP imaged at 30 weeks postmenstrual age (PMA) and no visible foveal depression, simulating pseudo hyperreflectivity in the inner retinal layers. B This pseudo-hyperreflectivity appears to migrate deeper within the inner retinal layers as the foveal depression appears and with better attenuation of the inner retinal layers in the foveal centre, as seen in the infant at 44 weeks PMA.

Fig. 2. Fallacies of imaging hyperreflectivity in the inner retina layers.

A Spectral-domain optical coherence tomography image was obtained from a preterm infant with Type 2 ROP, which subsequently resolved spontaneously. The infant was supine and imaged comfortably on the hand-held Bioptigen (Leica, USA). The ‘hyperreflectivity’ visible in the inner retina is commonly noted when there is non-fusion of the inner retinal layers and is visible within the layers between the outer plexiform and the internal limiting membrane, as seen in this image. B The ‘extent of this hyperreflectivity’ can vary based on the scan chosen for analysis, as evidenced by another image in the same session.

The authors have also reported that the presence of choroidal thinning before anti-VEGF is a biomarker for unfavourable outcomes after treatment. However, choroidal thickness in preterm infants is dependent on various factors. Firstly, choroid measurements are impacted by the positioning of the infant (supine vs. upright); an average measurement across the central 1 mm of the fovea may be more accurate. Secondly, the choroid increases in thickness linearly between 30 and 38 weeks postmenstrual age (PMA) [3]. The thickness Belenje et al. report, falls within the range of the average preterm, and they needed to adjust their findings with PMA.

The macular oedema they observed was first reported in Asian Indians with type 2 ROP was transient and resolved by 52 weeks PMA [4]. Optical coherence tomography angiography (OCTA) is useful for studying the superficial, deep capillary plexus and choroid in infants successfully treated for type 1 and 2 ROP. It may help determine the outcome after anti-VEGF better. We reported proangiogenic and inflammatory molecules in tears, which can be collected in a supine infant non-invasively and assessed for biomarkers that can predict disease progression and regression in treated or untreated ROP effectively [5].

Competing interests

The authors declare no competing interests.