ERM incidence in our control group (19.5%) at 24 months, this value is comparable to data available in the literature (2.2–26.1%), bearing in mind that patients’ mean age was 70, 29 ± 8,45, corresponding to the peak of ERM incidence reported by the Blue Mountains Eye Study between 74 and 85 years of age, with a stable prevalence in those older than 85 years [5, 6].

Ocular surgery plays an important role in ERM development. Patients undergone to cataract surgery reported an ERM incidence of 16.8% and following pars plana vitrectomy for rhegmatogenous retinal detachment ERM incidence ranged from 4.4 to 12.8% [7,8,9]. Nevertheless, literature about ERM incidence after glaucoma surgery is poor.

Vieria et al. reported an ERM incidence after trabeculectomy of 56%, of whom 9/50 eyes (18%) with PMF and 19/50 eyes (38%) with CMR. The mean follow-up time after surgery was 27.8 months [1].

In this study, the ERM incidence was 39.1% at 24 months, 12 eyes (29.3%) developed a CMR and 4 eyes (9.8%) a PMF. At 6 months after the surgery, ERM incidence was 29.3%; in particular, 12 (29.2%) eyes developed CMR and 2 (4.9%) PMF.

It is interestingly to note that ERM frequency increased, but CMR remained the most frequent ERM type with no alteration of macular profile. In the PMF cases, patients did not experienced vision loss or metamorphopsia at both time points during the follow-up.

Additional cataract surgery did not increase the ERM incidence after Ex-Press implant, because no significant difference was reported between ERM frequency in the combined surgery group and Ex-Press implant alone at each time points.

CFT and MV did not change significantly during the follow-up in the treated group. Also, no difference in terms of CFT and MV has been noted in the subgroup analysis between combined surgery and Ex-Press implant alone. Interestingly, CFT and MV remained comparable to reported average measures in healthy individuals ranging from 255.4 to 271.4 μm and from 6.76 to 8.53 mm3, respectively [10].

Furthermore, CFT and MV did not significantly increase between eyes with and without the ERM. This point confirms that most of ERM was CMR not altering the macular profile, and the 4 cases of PMF reported no visual loss or other symptoms.

Moreover, OCT is more sensible to detect ERM compared to biomicroscopy or fundus photography, increasing the incidence of ERM in patients where diagnosis is made with OCT. Therefore, ocular surgery is surely a risk factor developing ERM, but its impact should be overestimated, especially comparing the ERM frequency after surgery diagnosed with OCT, versus ERM incidence in healthy population, usually diagnosed with less sensitive methods.

In particular, a recent cross-sectional study on the same population of the Blue Mountains Eye Study reported ERM incidence of 21.4% using OCT for diagnosis, compared to the previous reported value of 7%, when biomicroscopy and stereo fundus photography had been used to make the diagnosis [5] [7]. Therefore, photographic diagnosis may underestimate ERM incidence by two- to threefold compared with OCT [5].

Furthermore, ERM was associated strongly with increasing age, a relationship that also has been reported in most other studies of ERM [11].This is likely related to the onset of PVD [12]. Cataract surgery is associated with ERM development, as already reported other studies, supporting the role of PVD in ERM development [6, 13, 14].

Nevertheless, the exact mechanism that leads to ERM formation is still debated, because the myofibroblastic pre-retinal cells are thought to transdifferentiate from glial and retinal pigment epithelial cells that reach the retinal surface via defects in the internal limiting membrane (ILM) or from the vitreous cavity [2].

Therefore, Ex-Press implant alone or combined with cataract surgery increases the incidence of ERM (39.1%), but less then trabeculectomy (56%) and more than cataract surgery alone (16.8%) [1, 7].

Therefore, we support the role of PVD in increasing ERM incidence, but Ex-Press implant surgery is less invasive compared to trabeculectomy, as demonstrated a minimal inflammation and scarring reactions after Ex-Press implant in rabbits [15]. The increased levels of postoperative inflammation and the rapid variation of IOP that occur after glaucoma surgery may induce or accelerate the PVD.

Nevertheless, Ex-Press implant procedure is less invasive compared to trabeculectomy, because iridectomy is not necessarily performed, maybe justifying the reduced ERM incidence in our study compared to data reported by Viera et al. [1].

Moreover, any role of postoperative hypotony was excluded because early postoperative IOP did not show a statistical difference between eyes that developed ERM and eyes without.

Also, the role of retinal break in ERM development after Ex-Press implant can be excluded, because during the follow-up period no treated eye experienced a retinal break or a retinal detachment.

A limitation of our study is the loss of 13 patients during the follow-up compared to our previous study population. Nevertheless, the calculation of sample size confirmed that a minimum of 38 patients was required to detect a 25.7% difference in the incidence of ERM, with a power of 80% and a significance level of 0.05.

Another limitation is the subjective evaluation of OCT images; however, each masked researcher analyzed all raw images of the macular area to detect any artifact, before evaluating the macular profile. In case of discordance between the masked researchers, the principal investigator reviewed and categorized the images.