Over time, BCVA of eyes with ODPM can deteriorate to 20/200 or worse in many cases; Of course, spontaneous remission has notably been reported in about 25% of cases (5, 6). With spontaneous remission, however, cystic changes of RPE and neurosensory retina and lamellar or full-thickness macular hole are likely to develop, with risk of permanent visual loss (5); serous macular detachment may also recur (7). Even in pediatric cases, whose chances for spontaneous resolution are higher (8), postponing the surgical interventions to monitor the natural course of maculopathy carries the risk of amblyopia development – resulting from visual impairment and changes in visual centers innervated by the affected ganglion cells (9). Hence, primary surgical management of ODPM is recommended, especially in cases with subretinal fluid (SRF) accumulation (1).
Pathology and pathogenesis in the literature
Historically, vitreous has been suggested as the primary source of fluid leakage into the retina through a thin porous membrane, constituted of dysplastic retinal tissue covering the ODP (10). Through serial histopathology sections of electron microscopy, Christoforidis et al. could visualize holes in the diaphanous membrane overlying the ODP bridging with a schisis-like cavity in the retina (11). SRF drainage through pores in the ODP roof was successfully performed by Johnson and Johnson (12) and Postel et al. (13). One hypothesis is that a negative pressure created by posterior hyaloid traction over the porous ODP-covering membrane generates an inward gradient of liquified vitreous into the retina (11, 14).
Although inconsistently, vitreous strands over the ONH and peripapillary retina have been detected on OCT scans and electron photomicrographs of patients with ODPM (2, 14). The role of vitreous traction in ODPM pathogenesis is amply supported by encouraging remission of maculopathy following vitrectomy (15) or spontaneous PVD (16). Strong vitreoretinal attachment at the disk margin has been reported during surgically-induced vitreous detachment (7). In addition to the role of posterior hyaloid traction, the tangential traction on the retina caused by ILM is also proposed to exert elevational traction on the retina, maintaining the inward fluid gradient into the retina (via ODP) (17). Post-vitrectomy subretinal migration of gas and silicone oil further supports the existence of communication between vitreous cavity and subretinal space through ODP; curiously, it happens when no vitreoretinal traction exists, indicating the involvement of other pathogenic mechanisms (12).
Cerebrospinal fluid (CSF) oozing from the adjacent subarachnoid space is the second plausible fluid source; it was first proposed in 1964 (18) after Regenbogen and colleagues noticed a pulsating transparent membrane over the ODP during surgery, which they attributed to CSF pressure fluctuations. Akiba et al. had a similar observation, which they ascribed to intravitreal traction caused by anomalous Cloquet's canal (19). Friberg and colleagues reported free pulsation of glial and vitreous remnant overlying ONH into and out of ODP in an eye with visible PVD (20). They also recognized a retrobulbar cyst communicating with the vitreous cavity through B-scan ultrasonography and documented a constant relative hypotony in the affected eye, presumably due to intraocular fluid drainage into the cyst and ultimately the subarachnoid space (20). In 2006, intracranial migration of silicone oil was noted on brain magnetic resonance imaging scans of a patient with ODPM presenting with a headache after vitrectomy and silicone oil injection (21). The direct communication of intraretinal cystoid spaces and the lamina cribrosa gap in the ODP (22) and their connectivity with the vitreous cavity were later visualized using high resolution, enhanced depth imaging (EDI) and swept-source (SS) OCT scans (23–25). As theorized by Johnson and Johnson (12), anomalous inter-connection and fluctuation of pressure gradient between intra- and extraocular spaces enable CSF and vitreous aqua to move through the ODP. When intracranial pressure (ICP) decreases, syneretic vitreous is sucked toward the communicating perineural space, and then with a rise in pressure, trapped vitreous and/or CSF is ejected into/under the adjacent retina and posterior vitreous cavity. The same mechanism could explain subretinal migration of gas and oil, post-vitrectomy, and occurrence of maculopathy in pediatric ages when liquified vitreous is uncommon (12).
As to the morphology and progression of ODPM, Lincoff et al. (26) first introduced the so-called bilaminar retinoschisis concept, in which liquid accumulation emanates from the inner neurosensory retina, extending outwards through outer layer lamellar holes. They proposed that true serous macular detachment occurs merely as a complication of longstanding intraretinal edema. Although this view has been widely accepted (23), it has been challenged by occasional observations of an outer layer hole in OCT scans (27, 28) and cases of ODPM with macular detachment, where no inner retinal schisis-like cavity was found (29). Todorich et al. reported a case of ODPM with direct connection of SRF to ODP, as detected on spectral-domain (SD)-OCT scans (30). Using high-resolution OCT, Imamura et al. showed that fluid could move straightly from ODP to multiple retinal layers, including the sub-ILM space, ganglion cell layer, inner and outer nuclear layers, and subretinal space (28). The outer nuclear layer is usually affected – as seen in our cases; one interpretation could be that in the majority of cases, the inner retina and/or subretinal space are involved secondarily to an initial passage of fluid through the outer retina (31). Intriguingly, Skaat et al. described two distinct OCT patterns in their cases: i) a predominant serous detachment pattern, with no to minimal schisis-changes of the photoreceptor layer in pediatric patients (mean age: 9 years old), versus ii) a multilayer schisis pattern in older adults (mean age: 31.7 years old) (32). The former pattern in younger patients has been reported by others as well, but it was not confined to the pediatric population (33). Although direct SRF conduit from ODP is rare (1), perhaps when present, it allows fluid passing beneath the retina much sooner than expected for conversion of asymptomatic ODP to ODPM in early adulthood. ODPM does occur unpredictably; however, blunt ocular and head trauma has been suggested as a potential trigger, especially in pediatric-onset cases; Rii et al. have attributed this to the severe hyaloid face adhesion in pediatric patients exerting an anteroposterior tractional pull on the macula following trauma (34). Another explanation could be the sudden rise in ICP after the trauma, forcing the CSF into/under the retina (3).
In summary, it could be reasonably inferred that both vitreous and CSF could serve as the pathogenic source of the accumulated fluid, passing into the retina through multiple layers, most commonly the outer nuclear layer (11). Moreover, both vitreoretinal traction and ICP pressure fluctuation play a role in the pathogenesis of the disease, either being prominent at certain ages and/or under different circumstances. We also suggest that a direct subretinal connection with the ODP cavity together with a traumatic experience may act as risk factors for accelerated progression toward ODPM in cases with asymptomatic ODP.
Surgical techniques in the literature
As previously mentioned, tractional forces over the ONH allow for the fluid entrance through the ODP, and the traction exerted upon the peripapillary and macular area could promote schisis separation of retinal layers, facilitating the fluid migration into the retina. Therefore, complete vitrectomy with PVD induction is the mainstay of ODPM treatment (3, 27). Hirakata et al. showed that complete retinal attachment in 7 of 8 eyes was achieved with isolated PPV and PVD induction, although it took up to a year (35). An alternative surgical approach is placing scleral buckles between the optic disc and macula, with a similar success rate of 85% as PPV and PVD induction. Besides alleviating vitreoretinal traction, scleral buckling is suggested to obstruct the fluid passage from ODP (36). However, this tricky technique is rarely applied in the management of ODPM (4).
ILM peeling, gas tamponade, and juxtapapillary endolaser photocoagulation are commonly utilized adjunctive procedures alongside PPV. The rationale behind these procedures and their additional benefit to vitrectomy is still controversial. Pneumatic retinopexy alleviates vitreomacular traction by inducing PVD and displacing accumulated fluid (37), but when applied without PPV, it has a temporary effect (38) and retinal reattachment rate of 50% with a mean number of 1.8 injections (39). Laser application is tentatively proposed to seal the route of the ODP to the fovea, but it has a very low success rate (40), probably because choroid and deep retinal layers absorb much of the laser energy. It also bears the risk of causing significant visual field defects (41). Combined gas tamponade and laser application had 75% success rate, but re-intervention was needed in 40% of the cases (42). After unsuccessful photocoagulation or pneumatic retinopexy, PPV and fluid air exchange, with or without laser treatment, have yielded an 88% reattachment rate (6). More recently, ILM peeling has been suggested to ensure the removal of all anteroposterior and surface-parallel tractional forces from the retina (43). Three multicenter studies evaluated the surgical success rates gained with PPV and PVD induction +/- ILM peeling, gas tamponade, and juxtapapillary endolaser and evaluated the extra therapeutic gain associated with each of these adjunctive treatments (1, 15, 44). They showed an overall 75–86% retinal attachment success rate, but none found a significant improvement gained with ILM peeling or temporal laser application; only gas tamponade showed additional efficiency in the study by Avci et al. (44); however, all these studies suffer from small sample sizes and heterogeneities in surgical procedures employed (1).
Marticorena et al. reported a case successfully treated with peeling of ILM after an initial failure of PVD and laser application (45). PVD induction, gas tamponade, and ILM peeling resulted in favorable outcomes in three ODPM cases within a few months, as described by Georgalas et al. (46). In a retrospective analysis of five patients who underwent PPV plus gas injection with or without ILM peeling, Skaat et al. reported complete SRF resolution in the former group whereas macular detachment persisted in patients for whom ILM peeling was not performed (32). Even considering the different clinical and morphological characteristics of the two groups, it could still be inferred that ILM peeling is a critical surgical maneuver for alleviation of maculopathy. Despite the formation of macular holes in 57% of the eyes operated, Shukla et al. achieved excellent results with vitrectomy, ILM peeling, and tamponing in 7 cases. 3 out of 4 holes were closed spontaneously, and the final visual outcome was unaffected by macular hole development during the recovery course (43). Fovea-sparing ILM peeling has been suggested to minimize the risk of macular hole formation (47); however, leaving a central island of ILM on the macula showed no protection against macular hole formation in one out of two eyes undergone the procedure (43). Moreover, full-thickness macular hole formation has been reported following PVD, gas tamponade, and laser photocoagulation without ILM peeling (48); as Shukla et al. proposed, most of the risk could be attributed to removing the strongly adherent posterior hyaloid face (43).
Spaide et al. introduced another adjunctive maneuver beside PPV, i.e., partial-thickness fenestration of the retina, radial to ONH. It is proposed to allow fluid redirection toward the vitreous cavity instead of the intra-/sub-retinal layers (49). They later achieved 94% foveal fluid resolution with this technique (50). However, a previous attempt to create a fenestration connected to schisis cavities by Slocumb and Johnson resulted in persistent macular detachment due to premature closure of the fenestration soon after surgery (51). Moreover, this technique will not be effective in the presence of a direct conduit beneath the retina.
Sealing the congenital pit with platelet-rich plasma (PRP) or fibrin glue has shown promising results and seems to shorten the long duration of restoration following surgery (30, 52). However, the plugs are temporary and cannot be regarded as a permeant solution. Long-term safety of
PRP is unknown as it can theoretically trigger proliferative vitreoretinopathy (30). The use of fibrin glue bears the risk of allergic reactions and microbial transmission (53).
Insertion of inverted ILM flap over the ODP has recently been suggested; the flap could act as a physiologic physical barrier against vitreous and oil migration and induce gliosis and cell proliferation within the ODP cavity (54). We observed pit remodeling and partial closure as soon as two months after complete PVD induction, ILM peeling, reverse ILM flap insertion, and gas tamponing. Moreover, we did not observe any macular hole formation in the three eyes operated, consistent with previous case reports using this technique (55, 56). Following a similar rationale, scleral autograft has also been proposed (57) but has an inherent risk of optic nerve damage (58).