What Africa Taught Me About Vitreoretinal Surgery: Part Two

It has been well documented that macular hole patients in Africa represent a rather unique patient demographic when compared to other parts of the world – they usually present late, and often with advanced cases of the disease. In fact, in our recent macular hole study (1) we observed that 54% of Black African patients presented with large macular holes with a median size of 500 microns. If this is compared to the UK BEAVRS (British and Eire Association of Vitreoretinal Surgeons) study, which reported a median of 394 microns, the difference is astounding.   

Less often discussed, however, is how the African eyes themselves can differ both in tissue characteristics and surgical response. From the little that has been published, posterior vitreous detachment (PVD) appears less common, vitreoretinal adhesion is stronger, and the internal limiting membrane (ILM) tends to be more adherent.

One of my earliest patients in Rwanda, who underwent epiretinal membrane (ERM) surgery with ILM removal as part of the procedure, returned with both visual improvement and restoration of the foveal contour. But on closer inspection, I noticed several small eccentric macular holes (EMHs) on the temporal side of the macula (Figure 1). At first I was intrigued by this finding, but then I let it be. And then the same phenomena began to appear in other patients too. 

To better understand this pattern, we reviewed 125 consecutive cases of macular surgery and found EMHs in 14 eyes, with an incidence rate of around 11.2%. On OCT scan, the majority of them were full thickness, although few were partial thickness as well. We later published these findings in Retina (2). The natural questions that arose from these findings were: Why do they form? Why are they always temporal? And why is there such a high incidence in this population when compared to patients of a predominantly Caucasian background?

First, let me clarify that these holes are not iatrogenic. They occurred across different surgeons, including myself and two visiting colleagues, and all cases involved dye-assisted peeling. The location was consistently temporal, not at the site where the peel was initiated, which is always superior. This consistent occurrence in the temporal macula may relate to its anatomical and vascular features. The temporal macula is thinner than the nasal macula, has smaller caliber end arterioles, and lies in a watershed zone between the retinal vascular arcades. These patients are also elderly, with vascular comorbidities such as hypertension, making chronic ischaemia another possible factor. Together, these features may increase susceptibility to ischemia and mechanical stress, favoring the development of EMH.

We postulated a few theories to explain why such a high incidence occurs in this population. First of all, with the ILM being more adherent in these eyes, there is a greater risk of Müller cell damage and apoptosis. ILM removal can damage Müller cells, with microscopic studies showing Müller cell processes attached to the removed ILM exhibiting necrosis and degeneration. In cases where the ILM is very adherent, as observed in our cohort, there may be increased damage to Müller cells with tearing out of their outer parts. Müller cells are responsible for the homeostatic and metabolic support of retinal neurons and the control of the extracellular space fluid.

Another possible cause of EMH is intraocular dye toxicity. At Kabgayi Eye Unit we use brilliant blue G, which appears less toxic than indocyanine green, though EMH has also been reported in cases without any dye use.

Another theory is postoperative tangential contracture and retraction of the ILM edge, which may cause the formation of EMH located at the edge of the peeled ILM. This would also explain the delay in their appearance. These mechanisms may complement each other, with ILM adherence, dye effects, and postoperative ILM contraction all contributing to the formation of EMH.

This experience changed my practice. I now no longer perform wide ILM peels. Once I achieve an adequate peel around the hole, I stop to minimize collateral damage. I also peel slowly with tangential force, aiming to avoid petechial haemorrhages on the macula, and sometimes interrupt the circular form of the ILM edge on the temporal side, since contraction of the edge may contribute to hole formation. Since adopting this approach, I have not observed the development of these EMHs.

It is worth noting that we have excellent macular hole surgery outcomes, with a closure rate of 93% (1). And these eccentric macular holes do not seem to cause any visual symptoms, and nor do they lead to retinal detachment.

One further point was raised by the editorial that accompanied our Retina paper (3) – the need for more research from low- and middle-income countries. Our findings highlight that surgical outcomes and complications can differ significantly in these settings, and that the global literature must begin to reflect this reality.

The lesson for me is simple – African eyes behave differently, and good surgeons need to be able to adapt to what they are confronted with, even if that means changing their long-standing techniques, to best benefit their patients.