Efficacy of Each Treatment on Each Etiology

The anti-VEGF seems to give interesting visual improvement, specially for ME following RVO.

Triamcinolone trend curves are more difficult to interpret since the treatment and the follow-up are interrupted shortly.

Vitrectomy and ILM peeling are very interesting, specially for DME and ME following RVO. The final improvement reaches 8 lines and the trend curve is rising. Surprisingly the lowest efficacy is for the one indisputable indication: the ERM.

Triamcinolone decreases the anti-VEGF efficacy; the final result is lower than that for anti-VEGF alone and the trend curves drop further with time.

The combination of ILM peeling and Triamcinolone provides much lower visual improvement than ILM peeling alone.

Efficacy of Each Treatment in General

In grouping the 3 ME etiologies together we come to the same conclusions.

Effects on Retinal Thickness

A reduction of macular thickness is achieved with all treatments. The maximum reduction is obtained after 9 to 12 months. Then the thickness of the retina seems to stabilize when anti-VEGF and ILM peeling are used. When Triamcinolone is used alone or in combination, the retinal thickness seems to increase again after one year.

We calculated the correlation coefficient between macular thickness and VA in order to obtain the linear degree of dependance between these two variables.This correlation coefficient is high (0.46)and significant before the treatment : the lower the VA, the thicker the macula.

When doing the same with post treatment VA, the correlation coefficient is much lower but still significant. This means that vision can go down even if the retina does not get thicker or, on the contrary, the vision can improve even if retinal thickness increases.

When comparing the two pre and post treatments coefficients, it appears that the link between VA and retina thickness is significantly much higher before the treatment than after.

In order to understand this phenomenon, we tried to analyze the difference between pre and post treatment macular thickness using box plots. This technique allows us to analyze a population by grouping 50% of the population in the box (separated in 2 by the median) and adding an indication of the upper and the lower desile between which 80% of the population is represented.

Here we see the distribution of the macular thickness populations before treatment.

And here after treatment. We can see that the population treated by ILM peeling has a smaller decrease and a smaller dispersion.

When considering the risk of development of atrophy, we see that all the groups, except the ILM peeling group, have at least 10% of their population with a thickness smaller than 180 microns. This can explain that a higher decrease in macular thickness is not necessarily linked with a better improvement.

Complications

The increase or the development of a new cataract is more frequently in the ILM group or the triamcinolone group and this is statistically significant.

IOP problems have been noticed more frequently in the Ozurdex group or the Triamcinolone group and this is statistically significant.

Surprisingly, no statistical difference has been observed as far as retinal detachment, choroidal detachment, vitreous hemorrhage or macular hole are concerned.

Finally

Finally, ILM peeling seems to provide the best results in all the ME etiologies without resulting in more complications than the other treatments.

This can be confirmed when comparing the improvements taking into consideration the initial VA.the silver medal goes to anti-VEGF and the bronze medal to the association ILM peeling plus Triamcinolone.

The lower improvements obtained in diabetic and in post RVO edemas could be explained by the lower percentage of ILM peeling made for these cases.