This is a narrative review conducted at the Ophthalmology Department of Fondazione Policlinico Universitario “A. Gemelli”, IRCCS. None of the writers performed any new investigations involving humans or animals; instead, this narrative review is solely based on earlier research, thus the Institutional Review Board approval was waived.

We searched the Cochrane Central, PubMed, Web of Science, and ClinicalTrials.gov databases for papers regarding EMAP, looking for the following keywords: “extensive macular atrophy with pseudodrusen”, “extensive macular atrophy with pseudodrusen-like appearance” and “EMAP”. Databases were accessed on May 2024. We only took into account English-language articles. None of the writers performed any new investigations involving humans or animals; instead, this narrative review is solely based on earlier research.

There is currently no documented etiology or established pathogenesis for EMAP. Risk factors for EMAP may not be the same as those for traditional AMD [10, 11]. Specifically, even if previous studies focused on small sample sizes, no apparent correlation was found with cardiovascular risk factors, such as smoking status, body mass index, hypertension, diabetes, low-density lipoprotein (LDL) cholesterol, serum total and high-density lipoprotein (HDL) cholesterol, and triglyceride levels [3]. These findings suggest that the development of EMAP is not related to cardiovascular imbalance or the ensuing vascular insufficiency at the choroid level, being significantly different from pseudodrusen AMD, which has previously been associated with cardiac risks, cardiovascular illness, and choriocapillaris perfusion imbalance [11, 12].

Conversely, the role of immunological pathways, particularly complement cascade, was investigated. The EMAP Case-control National Clinical Trial evaluated 115 patients from various French centers and found risk factors for complement pathway dysfunction (abnormal levels of C3 and CH50), gender (women tended to be more affected, with rates ranging from 52.5 to 60.8%), and other abnormal levels of inflammatory markers (erythrocyte sedimentation rate, eosinophils, and lymphocytes), as well as a potential correlation with a family history of glaucoma and AMD [3]. These findings support the theory that an inflammatory state with an autoimmune disease background serves as the trigger for EMAP pathogenesis [7]. Moreover, the development of RPD may the consequence of this inflammatory state, since in AMD the presence of RPD has been claimed as a biomarker of the retinal pigmented epithelium (RPE) inability to control and limit the damage related to the loss of regulatory immune responses (para-inflammation) [13].

As a result, similar to AMD, inflammation might play a role in EMAP, with further research needed to assess if complement cascade activation is involved to a greater extent [3]. Genetic analysis of risk alleles including those of the alternative complement pathway, was performed in 65 EMAP patients. Heterozygous and homozygous variations for the rs1061170 (His402Tyr) and for the rs800292 (Val62Ile) in CFH gene were reported, with the first being more frequent (around 80% of cases). Five and 25 patients were homozygous and heterozygous for the rs10490924 (Ala69Ser) in ARMS2 gene, respectively. For the rs2230199 (Arg102Gly) in C3, two patients were homozygous and 20 were heterozygous [2]. When comparing the rates of these variations with the genetic profile of classic AMD, no significant changes were reported regarding CFH and ARMS2 genes. Conversely, in EMAP patients, the mutation rs2230199 (Arg102Gly) in C3 was substantially more common [2].

Previous research linked the dysregulation of the alternative complement pathway to dense deposit disease (DDD), characterized by electron-dense deposits in the glomerular basement membrane (GBM) of the kidney, and basal laminar deposits (BLamD) in Bruch’s membrane (BrM) [14]. This term refers to the presence aberrant basement membrane proteins embedded in long-spacing collagen, situated between the natural basal lamina (BL) and the RPE plasma membrane [15]. Boon et al. reported the presence of BLamD in a variant of AMD with Tyr402His mutation in the CFH gene, featured by early onset and severe vision loss [16]. Similarly, Fragiotta et al. recently reported the presence of BLamD, located in the superior macula, in the very early stages of EMAP, suggesting a common pathway of immune dysregulation [15].

As of now, neither a definite genetic correlation, nor a familial transmission of the disease, have been demonstrated yet. Some cases of genetic correlation were recently reported, but a clear association with EMAP pathogenesis has yet to be defined [17, 18].

Douillard et al. recently carried out a second investigation with the same group of the EMAP Case-control National Clinical Trial, suggesting a direct neuronal degeneration involving rods and cones and identified a correlation between lifetime toxic exposure (pesticides used in agricultural operations) and EMAP [2]. Even though the precise molecular processes of neuronal degeneration are yet unknown, previous research highlighted that the majority of pesticides cause changes in the mitochondria, oxidative or endoplasmic reticulum stress, and ultimately death of neuronal cells [19]. Consequently, the authors suggested a possible parallelism between EMAP and other toxic-induced neuronal degeneration, such as Parkinson’s disease [2]. EMAP-induced neuronal degeneration may affect either rods and cones, but electrophysiology studies suggest a predominant cone apoptosis [2].

Even if only preliminary theories have been proposed, EMAP pathogenesis seems to be the result of a lifelong exposure to external toxic agents determining a state of subclinical chronic inflammation, which acts as a trigger in a substratum of pre-existing abnormal complement pathway regulation. Further studies are needed to give light to this hypothesis.

Since EMAP is a recently defined condition, a precise clinical course has not been defined, yet. Visual symptoms usually begin between 48 and 60 years of age, with a mean age of diagnosis of 55 years in a recent investigation [4]. Initially, patients refer difficulties with dark adaptation, which rapidly progress to the development of central scotomas. Inevitably, visual acuity deteriorates to legal blindness (vision of 20/200 or less) in 3 to 10 years from diagnosis [1]. In two studies analyzing visual outcomes in EMAP patients, best-corrected visual acuity (BCVA) declined linearly from baseline to the last visit, with a rate of around 7 Early Treatment Diabetic Retinopathy Study (ETDRS) letters lost per year [4, 20]. Overall, nearly 65% of eyes lost 3 or more ETDRS lines after 4 years of follow-up, with a BCVA of 35 ETDRS or less in 57% of eyes [4].

Nevertheless, the progression rate of BCVA, similar to that of RPE atrophy, showed significant interpatient variability, highlighting the need for further longitudinal studies [4].

Reticular pseudodrusen, also known as subretinal drusenoid deposits (SDDs), have been defined as well-defined conical (“dots”) or flat (“ribbon”) hyperreflective deposits among the RPE and the boundary between photoreceptor inner and outer segments [21]. In EMAP, several authors claimed the presence of RPD-like lesions as significant hallmark of the pathology.

In a group of EMAP patients ranging from 45 to 53 years, Fragiotta et al. reported the presence of RPD/SDD-like deposits without atrophic changes [15]. On structural OCT, SDD appeared in the superior perifovea and tended to coalescence horizontally into a flat, continuous, reflective material localized between the RPE and BrM, which determined a diffusely irregular RPE profile. The interdigitation zone (IZ) and ellipsoid zone (EZ) showed, in this area, a scattered granular hyperreflectivity [15]. Following a 4-year period of observation, an hyperreflective material had accumulated under the dysmorphic RPE and EZ/IZ lines were no more visible. The presence of this medium-reflectivity interposed substance determined a clear RPE–BL–BrM complex splitting with a noticeable BrM. At this point, complete RPE and outer retinal atrophy (cRORA) occurred, starting from the superior perifovea [15].

Similarly, Romano et al. reported a noticeable separation of the RPE from the BrM and thinning of the outer nuclear layer (ONL) on OCT, followed by the breakdown of the hyperreflective outer retinal bands, as the earliest findings of EMAP [20]. Among 17 eyes, RPE-BrM separation was already observed in all patients at presentation, when considering the extrafoveal macular areas, and in the majority of eyes (88.2%) at the foveal area. All eyes had overall thinning of the ONL, and 58.8% showed diffuse irregularity of the EZ, which became disrupted in seven eyes (41.2%) throughout the follow-up. In all instances, EZ disruption occurred prior to external limiting membrane (ELM) changes, and appeared in concomitance with RPE layer attenuation [20]. In the end, cRORA progressively developed with a typical barcode appearance, leaving space to persistent subretinal heterogeneous hyperreflective material and to subsidence of the outer plexiform layer (OPL). After an average of 17.5 months of follow-up, subfoveal fibrosis in the absence of intraretinal/subretinal fluid was seen in 11 eyes, starting in the perifoveal area and then spreading to the fovea [20] (Fig. 1).

At the ophthalmoscopic examination, EMAP’s atrophy assumes a predominant vertical orientation at the posterior pole. This feature derives from RPE changes primarily affecting superior perifoveal sector and then spreading towards the fovea. Pseudodrusen-like lesions and the early cRORA patches have a preferred superior localization probably due to the larger density of rod photoreceptors in those places as well as the lower perfusion caused by gravity [22, 23]. Unlike AMD, the majority of these RPD lesions have a “ribbon” appearance and uniformly extend to the retinal midperiphery rather than being restricted to the posterior pole. Conversely, “dot” pseudodrusen-like lesions are more uncommon in EMAP patients and are restricted to the macular area, similar to AMD [24]. When macular atrophy appears, it shows an oval shape progressively evolving from the superior macula to the other Early Treatment Diabetic Retinopathy Study (ETDRS) sectors [20].

Only few reports have been presented regarding the application of OCT angiography (OCTA) technology in EMAP patients. Kovach was the first to analyze OCTA findings in a patient affected by EMAP, reporting marked absence of choriocapillaris (CC) flow and mild attenuation of retinal vasculature [5].

Successively, Rajabian et al. performed an OCTA investigation comparing 14 eyes with EMAP with healthy controls. With respect to the progression of the disease, they divided the macular area in an atrophic area, a junctional area (which became atrophic during the follow-up) and a preserved area (non-atrophic) [6]. OCTA showed an almost preserved superficial capillary plexus (SCP) in all zones, when compared with controls. Conversely, they showed a reduction of vessel density in the deep capillary plexus (DCP) in either the junctional and atrophic zones, but not in the preserved zone, suggesting that DCP changes appear earlier in EMAP [6]. Interestingly, CC vascular architecture showed progressive damage when going from the preserved retina (highest vessel density), to the junctional region (intermediate vessel density) and to the atrophic region (lowest vessel density), with statistically significant differences between the three regions [6]. In conclusion, the authors hypothesized that CC alterations may occur more slowly in EMAP physiophatology, rather than DCP alterations which were most clearly identifiable in the junctional area [12].

Blue-light fundus autofluorescence (BAF) has been used to assess the progression of macular atrophy in EMAP. In a study carried out by Romano et al. on 36 eyes, BAF showed signs of RPE atrophy at the posterior pole in all instances, with 72.2% of eyes demonstrating hyperautofluorescent borders, whereas the remaining 27.8% showed iso-autofluorescent borders [4]. They reported that the advancement rate in EMAP is greatly impacted by the identification of iso-autofluorescent and more irregular atrophy boundaries. While the irregular atrophy border structure has been associated with a quicker rate of atrophy development in other maculopathies, the autofluorescence patterns found in the junctional area appeared different from classic geographic atrophy (GA) [4]. However, although the identification of hyper-autofluorescent borders was initially linked to rapidly developing forms of GA, recent clinicopathologic investigations detected that the hyper-autofluorescent rim is probably caused by stacked RPE cells and not associated with atrophy enlargement rate [25, 26]. Romano et al. hypothesized that atrophy will most likely enlarge in locations where RPD deposits have coalesced, and that the visible iso-autofluorescent pattern is the result of an heterogeneous material composition in the junctional area [4] (Fig. 2).

The annual rate of atrophy expansion, evaluated with BAF, was 2.9 mm2/year, leading to foveal involvement in 83% of eyes at the end of follow-up. This progression rate was higher than AMD-related GA progression (range 0.53–2.60 mm2/year), but more similar to the diffuse-trickling geographic atrophy (DTGA) subtype [27].

Fragiotta et al. reported, in the early phases of the pathology, a diffusely reduced BAF signal in the foveal region corresponding to RPE rarefaction with a remarkable RPE-BrM splitting and sub-RPE basal laminar deposits (BLamD) accumulation [15]. This reduced signal may be the effect of an altered metabolic trafficking between choroid and RPE, leading to a local deficiency of vitamin A, or may be caused by the loss or rearrangement of autofluorescent organelles from individual RPE cells [28, 29]. Furthermore, a fine granular hyperautofluorescence was noted at the boundaries of the atrophic lesion, probably originating from the staining of the sub-RPE material [15] [Figure 1].

Courtesy of Fragiotta et al. [15] Antropoli et al. [41] and Romano et al. [4]

Optical coherence tomography B-scans in extensive macular atrophy with pseudodrusen (EMAP), at different stages of evolution. A Early findings show an initial retinal pigmented epithelium-basal lamina-Bruch’s membrane (RPE-BL-BrM) complex splitting (white arrowheads) with an overlying dysmorphic and irregular RPE in the superior perifovea. B Subretinal deposits (SDDs) may show a “ribbon” shape (yellow arrow) or a “dot” shape (turquoise arrow). C In intermediate stages, RPE-BL-BrM splitting is visible in the foveal area, with hyperreflective material under the RPE. Moreover, a localized area of complete RPE and outer retinal atrophy (cRORA) is visible (yellow arrowhead); D Macular scan showing collapse of the nasal perifoveal outer retinal layers with choroidal hypertransmission (yellow arrowheads), while the fovea is still unaffected and shows RPE irregularities; E In late stages of the disease, cRORA involves the foveal area and subretinal fibrosis may develop (red stars). F Complete choroidal hypertransmission with clear visibility of choroidal vessels (white arrows) due to complete RPE atrophy.

Rabinovich et al. exploited the the semi-automated RegionFinder™ software (Heidelberg, Germany) to assess the development of atrophy in eyes with EMAP, compared to eyes with a subtype of Stargardt disease, known as fundus flavimaculatus (FFM): the first appeared to be about six times quicker than the latter (3.73 vs. 0.70 mm2/year) [8]. Atrophic expansion linked to both FFM and EMAP was non-linear over time, since the progression rate slowed down after an average of 1.3–1.6 years [8].

Recently, Battaglia Parodi et al. described peripheral retina alterations in EMAP patients using ultrawidefield fundus photography (UWFFP) and ultrawidefield fundus autofluorescence (UWF-FAF). They identified three main types of peripheral degeneration: RPE alterations, pavingstone-like changes, and pigmented chorioretinal atrophy. These degenerations progressed in two-third of eyes over time, at a median rate of 0.7 sectors/year. Moreover, enlarging rate of macular atrophy, after the square root transformation, was 0.46 ± 0.28 mm/year [30].

A recent research proposed a three-stage EMAP categorization scheme, based on multimodal imaging results and visual symptoms [20] (Table 1). Moreover, a recent report distinguished EMAP pattern based on topographic involvement: predominantly central, predominantly peripheral and mixed forms [30].

Stage 1 is characterized by the splitting of RPE–BrM and scattered RPDs extending into the retinal midperiphery, with non-confluent patches of RPE atrophy limited to the superior macula. In this stage, a good preservation of visual acuity is maintained. In Stage 2, RPE atrophy appears confluent and widespread, acquiring a prominent vertical direction, but the fovea is still preserved. Though visual acuity stays comparatively high, visual symptoms such as reading difficulties and poor dark adaption become more noticeable. In Stage 3, the illness affects the fovea, with a consequent significant decline in visual acuity. Moreover, foveal involvement may result from either RPE atrophy (stage 3 A) or subretinal fibrosis development (stage 3B) [20] (Fig. 3).

Blue fundus autofluorescence photographs of extensive macular atrophy with pseudodrusen (EMAP), showing the classic evolution pattern. Hypoautofluorescence lobular lesions are visible in perifoveal areas, particularly the superior perifovea (A). These multifocal lesions tend to coalesce into a single patch of dark and sharply demarcated atrophy, initially not affecting the fovea (B). In advanced stages, a vertically oriented area of atrophy involves the macular region and involves the fovea (C). Note the hyperautofluorescent rim visible at the boundaries of the atrophic lesion. Courtesy of Antropoli et al. [41], Romano et al. [4] and Vilela et al. [9]

Courtesy of Fragiotta et al. [15] and Romano et al. [20]

Staging system proposed by Romano et al., with the addition of the pre-extensive macular atrophy with pseudodrusen (pre-EMAP) stage, proposed by Fragiotta et al. Mild autofluorescence (A) and infrared (B) alterations ar