Regarding PHL, 86.4% of the eyes with pre-operative PHL (either focal or widespread) showed intra-operative evidence of mVCRs. The presence of PHL on pre-operative OCT resulted to be highly sensitive (98.1%) but poorly specific (42.9%) for the intra-operative detection of mVCRs. Multivariate analysis only confirmed a significant association between widespread PHL and mVCRs and postoperative OCT imaging showed residual traces of PHL in 25 out of 40 eyes that had undergone mVCR peeling. It is possible that what is appreciated on OCT as PHL is made of several layers, and during the operation we successfully removed only the most superficial (what appeared intra-operatively as mVCRs) leaving behind the outermost layer. Among the outermost layers, there could also be the ILM in cases in which strong adhesions exist between VCRs and ILM. Actually, Cho et al.
8 reported a 30% incidence of postoperative ERM during a follow-up period of at least 6 months in eyes with RRD that had undergone PPV and scraping of the VCRs at the macula. This figure contrasts with the incidence of postoperative ERM after ILM peeling during PPV for RRD that has been reported to range between 0% and 9% in a follow-up period of 6 to 12 months
25–27 and suggest that, although the surgeon may have the impression of thoroughly scraping away VCRs from the macula during the operation, some residual may be inadvertently left on the retinal surface and promoted subsequent membranes growth. Interestingly, a landmark histopathologic study on preretinal membranes,
28 revealed that in eyes with vitreous detachment and with residual macular “cortical vitreous membranes,” those membranes are often multilaminated and comprised of alternating layers of glial cells and cortical vitreous. More recently, a histopathological analysis of surgical specimens harvested from eyes with primary RRD identified different compositions of VCRs/membranes in terms of collagen and cells that were attributed to different stages in the PVR formation.
29 Thus, it is possible that, depending on a specific histological composition of VCRs, their removal may result more difficult and incomplete in some cases. As a consequence, if a correspondence really exists between PHL and VCRs, we expect to find in some eyes postoperative PHL persistency despite an apparently successful intra-operative VCRs removal. In these eyes, only the peeling of ILM could guarantee a complete VCR removal. Indeed, we observed that in no case in which the ILM was inadvertently peeled off along with VCRs, the PHL was still visible after the operation. Alternatively, it is possible that, according to a model previously theorized for the pathogenesis of macular pucker,
4 the scraping/peeling maneuvers to remove VCRs create microbreaks in the ILM through which glial cells may migrate and proliferate giving rise to the PHL. However, this hypothesis seems to be unlikely, because PHL was visible since the first month after the operation.