Structural changes of the macula and optic nerve head in the remaining eyes after enucleation for retinoblastoma: an optical coherence tomography study

Retinoblastoma (RB) is the most common type of primary intraocular cancer in children, representing 4% of all pediatric malignancies [1]. It typically presents at a young age, with about two - thirds of all patients diagnosed before 2 years of age [2].

RB1 gene is the first cloned tumor suppressor gene; the protein encoded by this gene is RB protein (RBp) which acts as a cell cycle regulator in the retina and many other tissues [3]. It is a part of a family of proteins including p 103 and p 107, known collectively as ‘pocket proteins’ which induce G1 arrest [4, 5]. RBp influences cell division, survival, and differentiation, but its role in the retina and consequences of its loss on retinal development and differentiation is still unclear and under research. RBp is detectable in the embryonic ganglion cell layer (GCL) but not the primitive retinal neuroepithelium, suggesting that it may regulate cell birth and or differentiation [6, 7]. RB1 inactivation leads to epigenetic changes in key cancer and differentiation pathways in the developing retina [8, 9].

The 5 - year survival rate of more than 95% of all children with RB is the highest of all pediatric cancers in developing countries [1]. Considering the high survival rate of RB patients and the severe impact of the late effects of RB and its treatment, it is important to evaluate health - related quality of life of RB survivors specifically vision related aspects that may result in disabilities, and subsequent restrictions in the activities of daily living which is less well known [10].

Effects of monocular enucleation on ipsilateral and contralateral projections in the higher centers were thoroughly investigated experimentally in different animal species [11‐16] long time ago. Objective assessment of the structural changes of the macula and optic nerve head of the remaining free eye after monocular enucleation in human didn’t yet discussed.

New imaging technologies for the early detection and objective evaluation of structural damage of the retina and optic nerve have been developed even before functional damage. Optical coherence tomography (OCT) provides real - time, objective, and reproducible measurements of the optic nerve head and retinal nerve fiber layer [17].

The aim of this study was to describe objectively the possible structural changes of the macula and optic nerve head in the free eyes of unilaterally cured retinoblastoma patients and to compare them with the changes that could be detected in the remaining eyes after enucleation using spectral domain optical coherence tomography.

The study included 41 (68.3%) males and 19 (31.7%) females with a mean age of 11.95 years ±4.0 SD (median: 12, range: 5–19). Mean BCVA (Log.MAR) in the studied eyes was 0.1 ± 0.03 SD (median: 0.1, range: 0.1–0.3) and mean spherical equivalent of 1.0 diopters ±0.73 SD (median: 1.0, range: −1.25 to +2.75). Mean IOP was 11.67 mmHg ±1.7 SD (median 12, range: 9–15). Average SAP mean deviation of −1.44 dB ± 1.48 SD (median: −1.19, range: - 2.01 to – 4.29) and pattern standard deviation of 1.14 dB ± 1.03 SD (median: 1.14, range: 0.0–3.1). The mean age at time of presentation of RB was 23.13 months ±13.36 SD (median: 24, range: 2.0–48) and 15.20 months ±7.76 SD (median: 15, range: 5.0–26.0) in study groups (I) and (II) respectively. The mean duration of follow up of RB children was 7.93 years ±3.49 SD (median: 7.5, range: 3.5–17) and 10.38 years ±3.73 SD (median: 8.4, range: 7.0–17.5) in study groups (I) and (II) respectively. There was no statistically significant difference between the control and study groups regarding all demographic data (P- values were >0.05). Table 1 summarized demographic data of the control group and study groups (I) and (II).

Table 2 demonstrated the comparison between the control and study groups regarding total and inner retinal layers thicknesses at the macula.

The mean GCL was significantly lower in study group (II) compared to each of the control group and study group (I). The superior sector in the 3 mm ring was the most thinner statistically in study group (II) when compared to the control group, and the nasal sector in the 6 mm ring was most thinner statistically in study group (II) when compared to the study group (I).

Comparison between the three groups revealed a statistically significant lower value in the overall GCC in study group (II) (P- value was 0.002). Comparison was done between the three groups regarding the mean GCC in all sectors show a statistically significant difference in nasal and temporal sectors in inner 3 mm ring and 6 mm ring (P-values were, 0.004, 0.017, < 0.01 and 0.042 respectively). Also a statistically difference was found in superior sector in 3 mm ring and inferior sector in 6 mm ring (P-value was 0.02 using Kruskal-Wallis test and was 0.05 using one – way Anova test respectively).

The mean GCC thickness was statistically significant lower in study group (II) compared to each of the control group and study group (I) with affection of all sectors except the inferior sectors which were statistically preserved.

The mean ONL thickness was preserved in all sectors with no statistically significant difference between the control group and the study groups except in superior, nasal and inferior sectors which were statistically lower in study group (II) compared to control group and study group (I). Table 3 provided the comparison between the control and study groups regarding inner nuclear, outer nuclear and RPE- IS/OS layer thicknesses at the macula.

A statistically significant difference was found in superior sector in 6 mm ring when comparison was done between the three groups using Kruskal-Wallis test. The mean RPE- IS/OS layer was statistically significant thinner in superior and inferior sectors in 6 mm ring in study group (II) compared to control group and study group (I).

The mean pRNFL thickness was statistically thicker in the study group (I) when compared to control group, All optic nerve head quadrants were statistically significantly thicker except the nasal quadrant. There was a statistically significant thin average pRNFL thickness, thickness of superior, temporal and inferior quadrants in study group (II) compared to study group (I).

Table 4 show comparison between the control and study groups as regards the average pRNFL thickness and the other optic disc parameters.

Figures 3 and 4a, e demonstrated retinal layers thicknesses at the macula and pRNFL thickness and optic nerve head parameters in one of our patients in study group (I) and (II) respectively. Pearson’s correlation coefficient was done between average macular layers thicknesses and average pRNFL thickness and different variables: mean age, sex, family history, consanguinity, duration of follow up, BCVA, spherical equivalent, IOP, mean SAP mean deviation and pattern standard deviation revealed no statistically significant correlations (P-values were >0.05) except a statistically significant positive moderate correlation in the enucleation group [study group (I)] between mean pRNFL thickness and mean age at the time of the study and mean SAP pattern standard deviation (r:0.65, P-value was 0.02 and r: 0.67, P-value was 0.02 respectively).

In the current study, thinning of retinal layers in the free eyes of the RB survivors was detected compared to that in the control group. This was statistically significant in the mean CFT, GCL, GCC, total macular thicknesses and some sectors of ONL. This was combined with preservation of the mean thicknesses of almost all RPE IS/OS layer, pRNFL and the other optic nerve head parameters. All those RB patients had normal vision according to World Health Organization guidelines [19]. The present study was the first study to report these findings. Indeed, to understand why this thinning occurred we reviewed how the retina developed and the role of RBp during cell cycle, division and differentiation and the effects of its loss on the retinal development beside RB tumor formation in certain locations.

All studies on this context was in animal models, two studies [20, 21] show that in the absence of RBp, the mature retina lacks ganglion and bipolar cells and has fewer rods and in the absence of both RBp and p107; ganglion, bipolar, rod and cone cells are all deleted. However, amacrine, horizontal and Muller cells survive either RBp or RBp and p107 loss.

In addition to the known role of RBp in controlling cell division, it binds and potentiates several transcription factors involved in differentiation [22, 23].

RB - deficient retinas had ectopic S - phase and high levels of p53 - independent apoptosis, especially in the differentiating retinal GCL in late embryogenesis and adults showed loss of photoreceptors and bipolar cells [21].

On the other hand, Vooijs et al. [24] in their study suggested that RBp, p107 and p53 are not required for terminal differentiation of photoreceptors. Continued proliferation of photoreceptor precursors or phenotype plasticity of the neuronal and glial cells reduced the effects of RBp deficiency in the neural retina [25].

Brg1 (Smarca4) is an ATPase subunit and a tumor suppressor in RB, It was found to regulate retinal size by controlling cell cycle length, cell cycle exit and cell survival during development. The combination of defective cell differentiation and lamination led to retinal hypocellularity in all three cellular layers of the retina and degeneration in Brg1-deficient retinae [26]. So RBp and Brg1 defects may be considered as a cause of CFT, GCL, GCC layers thinning in the maculae of free eyes in study group (II).

Cyclins and cyclin dependent kinases are direct regulators of RB family. If persistent expression of cyclin D1 occurred due to genetic defects, it disrupts photoreceptor differentiation as manifested by delayed expression of opsin in developing photoreceptors and alters their number and morphology in mature retina. These alterations were accompanied by disorganization of inner nuclear and plexiform layers [27]. It may have a role in genetically predisposed thinned retina in RB patients in the present study.

It was suggested that viral oncoproteins may have a role in retinal development beside predisposition to tumorogenesis [28‐31].

Another issue is the environmental signals to retinal cells by changing the intrinsic or extrinsic factors [32, 33].

Following monocular enucleation of RB patients in the present study, the contralateral clinically free eyes show thickened total macula significantly except in inferior sectors, the average GCL, GCC, and mean pRNFL as a whole and in quadrants of the optic nerve head except the nasal quadrant compared to same parameters in the contralateral clinically free eyes of RB patients not treated by enucleation. No statistically significant difference between study group (I) and the control group. Our finding here was that the thinning observed in study group (II) was overcome by thickening in study group (II) following enucleation to reach the normal thickness as in the control group. A statistically non significant difference between the study groups regarding the demographic and clinical data allowed us to do such a comparison.

The explanation of the plastic changes in the macula and optic nerve head of the remaining eyes after unilateral enucleation for RB was suggested to be actual increase in number of the ganglion cells and glial elements in an already thinned RB retina. A statistically significant increase in mean pRNFL in superior, temporal and inferior quadrants was detected due to extensive sprouting of nerve endings. We think that the increase in mean GCL and mean GCC were not significant because of associated decrease of the contralaterally projecting cells of the retina.

The net result between the actual increase in ipsilateral projection and the decrease in number of contralateral projecting cells in the retina was not significant. We couldn’t judge if there was a change in certain type of ganglion cells either in number and size, these need further histopathological studies. There was no explanation for more significant reduction in mean CFT after unilateral enucleation except that these increase in number of cells and glial elements involving more the inner retinal layers which are normally deficient in the fovea pit so the mean CFT here was still thin as found in study group (II).

These findings in macula and optic nerve head following unilateral enucleation in the present study can be explained by several animal studies [34‐37]. They found a significant increase in the number of optic nerve axons occurs in dorsal lateral geniculate nucleus and in the superior colliculus on the side ipsilateral to the remaining eye [35, 37‐39]. This apparent ipsilateral expansion may be explained in various ways; there might be no actual increase in the number of ipsilaterally projecting ganglion cells and the expansion may be due to a profuse sprouting of their axon terminals [36].

Another opinion was the rerouting of some retinal ganglion cell axons from the contralateral to the ipsilateral optic tract occurred without increase in ganglion cells number [36, 40] . Some authors suggested that neonatal monocular enucleation suppressed the otherwise occurring cell death of the ipsilaterally projecting ganglion cells [35] so that the latter appeared increased compared with those in normal rats [41].

More densely packed distribution of ganglion cells in lower temporal crescent region of the retina was found following monocular enucleation than in normal rats [42]. The second finding was expanded distribution of these cells in the upper hemiretina and lower nasal retina. This was in conjunction with their ealier results [43, 44]. A similar expansion of ipsilaterally projected ganglion cells into the upper nasal retina reported for retinocollicular pathway [39].

The third finding was decreased density of contralaterally projecting ganglion cells in retina compared with normal control. They found that some ganglion cells in the temporal retina have changed their axonal projection from the contralateral to the ipsilateral hemisphere and large to medium size cells lost in contralateral projection and gained in ipsilateral one [42]. This last finding about the type of ganglion cells involved in expansion was supported by Shirokawa et al. [45].

Lund et al. [46] proposed that some of the optic fibers which normally innervate the contralateral visual centers come to innervate the ipsilateral side, resulting in an increase in the uncrossed projections.

Ahmed et al. [47] proposed that the mechanism of persistence of bilaterally branched axons present early in the development. They found that initially the double-labeled cells are distributed over the entire retina but that over the course of development they are subsequently found in the ventral - temporal crescent in the normal adult albinos due to the withdrawal of axons belonging to retinal ganglion cells which die off.

One of the chemical factors which involved in neural plasticity after monocular deprivation is Retinal Brain Derived Neurotrophic Factor [48]. This may highlight the possibility of presence of chemical agents involved in the results of the effects of monocular enucleation on the contralateral eye which are not studied yet.

After unilateral enucleation in hamster at birth, examination of adult retinal ganglion cell layers (including retinal ganglion cells, displaced amacrine cells and glial cells [49]) in remaining eyes revealed increase by about 8%, with different distribution of cells according to cell size. The increase in cell number was found across the entire retina, but was largest in the temporal retina. [50]. The soma size is controlled by one factor which is the volume of axon of it had to maintain [39] also presence of overlap and competition increased terminal arborization and soma size [50].

There was a significant positive correlation between SAP pattern deviations and the pRNFL thickness in study group (I); increased thickness may be associated with further functional damage. So those survived RB children need a long term follow up by visual field for the remaining eyes. A study was conducted to evaluate the impact of RB on the health status of survivors in terms of disabilities and worries, both of which may restrict participation in activities of daily life. Fear of developing second primary tumors passing RB on to the next generation and of further loss of vision were important life-long problems they worried about [10].

This study was the first one considering description of possible structural changes in the macula and optic nerve head of the clinically free eyes of RB survived patients. These eyes were characterized by thinning of CFT, GCL, GCC in the macula with no statistically significant difference in thickness of pRNFL compared to the control group. These changes may be due to the genetic defects in RB. After unilateral enucleation, increased thickness in macular GCC and pRNFL were detected compared to RB group and these changes were statistically correlated to SAP pattern deviations. This may highlight the importance of long term follow up examinations of those remaining eyes. Further histopathological studies are needed to confirm these findings.