Retinoblastoma – post 2

I’m going to focus on the optic nerve in this post, starting with an example of normal histology and then touching on a non-cancer abnormality before covering the different grades of retinoblastoma involvement.

Ideally, when grossing a retinoblastoma globe, you should take a transverse section of the optic nerve. However, take care that you are indeed transecting the nerve (and not just the dural sheath), and be careful that you’re not cutting it so close to the globe that you cut off the lamina cribrosa or buttonhole the eye.

We had some debate at the LVPEI visit kindly arranged by Dr Ruchi Mittal, as to whether to embed the transverse section on the surgeon’s cut surface (i.e. closer to the chiasm, and the “true” resection margin) or on the pathologist’s cut surface (closer to the globe, and not representing the “true” resection margin). As with many grossing decisions, there are pros and cons for each approach. My general advice would be to decide on one or other approach (preferably consistently), and to be aware of how your chosen approach may impact your reporting. I’d welcome your comments on your selected approach.

---

Normal optic nerve

Here’s a low power longitudinal section of an enucleation specimen including the optic nerve head and central retinal vein. I’ve highlighted the area of the lamina cribrosa.

This is a higher power view of the same section, where the collagen fibres making up the meshwork of the lamina cribrosa can be better appreciated.

This is a transverse section of an optic nerve which is slightly atrophic. The white space between the dura and nerve may be artefactual, but on the higher powers the deeply eosinophilic fibrous septa are more prominent than I would expect.

---

Cupped optic nerve

Cupping is an abnormality of the optic nerve head which may complicate assessment of retinoblastoma, as it distorts the anatomy. It is often due to glaucoma.

This low power longitudinal view shows excavation of the ONH, with an overhanging edge.

The location of the lamina cribrosa is indistinct.

I’ve marked the lamina cribrosa’s rough position, but it’s not as clear as with the normal optic nerve head.

Practical point: If you are grossing a globe with extreme disc cupping and cut the optic nerve transverse section too close to the eye, it is possible to “buttonhole” the specimen.

---

Retinoblastoma involvement of optic nerve

In this example, the tumour overlies the optic nerve head but does not involve it.

Here, tumour extends into the superficial part of the ONH, but not quite into the lamina cribrosa. This is Grade 1 involvement.

Here, tumour extends into, but not beyond, the lamina cribrosa. This is Grade 2 involvement. Because of the discontinuous nature of the lamina cribrosa, assessing the extent of tumour infiltration may not be clear cut.

This is extensive tumour infiltration of the optic nerve which extends all the way to the cut end. Macroscopically, this globe was submitted with 2mm of optic nerve. This is Grade 4 involvement.

This is a transverse section of optic nerve infiltrated by tumour, so either Grade 3 or Grade 4 involvement, depending on whether the tumour extends to the resection margin.

Please let me know if you have any comments or questions about the optic nerve, particularly in relation to retinoblastoma. In my next post I will focus on the choroid.