Hyperreflective material in patients with non-neovascular pachychoroid disease

Our observational case series demonstrated that there are some cases of non-neovascular disease with HRM. All cases had thickened choroids and pachyvessels at a relatively young age. Furthermore, HRM improved spontaneously without intervention in nine eyes (81.8%) and finally resulted in so-called PPE or FCE. Although these cases are very rare in a real-world clinical setting, it could be a new entity of pachychoroid spectrum disease or an early stage of PPE or FCE. In addition, these cases might be misdiagnosed as AMD or ICNV because of the existence of HRM and suspected MNV. However, it is important that these cases should be diagnosed as non-neovascular disease using multimodal imaging and carefully observed without any intervention. To our knowledge, non-neovascular pachychoroid disorder with HRM reported in this study has not previously been described.

In the previous studies, HRM at the subretinal space were reported as subretinal hyperreflective material (SHRM). In the Comparison of Age-Related Macular degeneration Treatments Trials (CATT), SHRM was an important morphologic biomarker for worse visual acuity in patients with AMD [12]. SHRM is known to have several components, such as neovascular tissue, fibrosis, hemorrhage, and exudation, likely comprising serum, fibrin, and inflammatory cells [13‐15]. However, in our cases, it is deniable that HRM consists of neovascular tissue, fibrosis, or hemorrhage based on fundoscopic findings or multimodal imaging. Therefore, in this study, it is highly possible that HRM components were some kind of exudation. A previous study reported that in patients with CSC, which is categorized as one type of pachychoroid spectrum disease, breaches in the RPE could allow a large molecule, such as fibrin, to enter from the sub-RPE into the subretinal spaces [16]. Although the reason why patients in our study demonstrated HRM is uncertain, damage to the choriocapillaris, which was caused by the pachyvessels [17], would allow a large molecule such as fibrin to escape into the sub-RPE space. Furthermore, damage to the RPE would allow this molecule to enter from the sub-RPE to the subretinal spaces or neurosensory retina.

In our study, HRM were seen not only at the subretinal space but also at the OPL and ONL. Previous studies described that hyperreflective foci (HRF) [18], which is defined as hyperreflective speckling in the neurosensory retina detected by using SD-OCT, have been identified in various chorioretinal diseases [19‐21]. In this study, 4 eyes (case 1, 4, 5, and 6) had HRF, and these cases had acoustic hyperreflectivity behind the HRM. Proposed origin of HRF have several hypotheses, such as displacing RPE cells [18], activated microglia cells [22], and lipid extravasation [23]. Therefore, in some cases of this study, HRM might consist of these components.

In nine cases (81.8%), HRM improved spontaneously and four of them resulted in so-called PPE and three of them showed FCE. Symptoms such as metamorphopsia or distortion also improved in these eyes where the HRM ameliorated. Because both PPE and FCE are usually asymptomatic, we speculate that HRM was the cause of their symptoms. A previous study reported the morphologic characteristics of patients with CSC evaluated by SD-OCT [24]. Those authors described that SRF with subfoveal fibrin was more commonly observed in the early period of acute CSC than in eyes with chronic CSC. We speculate that having HRM indicates an active stage, and these characteristics disappear spontaneously when the activity declines. In addition, RPE alterations occur after improvement of HRM. Monés et al. described that some patients with soft drusen due to AMD showed isoreflective dots at the outer retinal layers associated with RPE defects. They have termed this ‘drusen ooze’, which was associated with a high risk of developing atrophy [25]. They hypothesized that drusen ooze could activate the surface of RPE, resulting in an increase in phagocytosis of extracellular debris that exceeds the RPE capacity, and finally leads to RPE death. Morphologic characteristics in our study are similar to that of their study. Therefore, HRM may activate the RPE apical surfaces, phagocytosed by RPE cells, and finally morphologic features might be similar to PPE.

Furthermore, three eyes (27.3%) showed FCE after the resolution of HRM. Although the pathogenesis of FCE formation is still uncertain, previous reports described that FCE lesions are congenital choroidal abnormalities [9, 26]. On the other hand, some reports mentioned that FCE may occur in associated with inflammatory diseases such as multiple evanescent white dot syndrome, multifocal choroiditis, or Vogt–Koyanagi–Harada disease [27‐29]. Similar to these cases, HRM in the neurosensory retina, due to inflammatory process, could be a mechanism leading to FCE formation. However, all cases in our study didn’t need anti-inflammatory therapy and improved spontaneously. Inflammation caused by HRM may be weak enough not to require treatment. Otherwise, it is a self-limited disorder.

In this study, the differential diagnosis was neovascular diseases, such as ICNV and AMD, because SHRM, which is derived from exudation, is generally associated with active MNV and often regresses as the MNV becomes quiescent after anti-VEGF treatment [14, 30]. However, all cases in this study had no MNV by multimodal imaging and were finally observed carefully without any intervention. Because the current optimal treatment for either ICNV or AMD is anti-VEGF agents [31, 32], it is important to check the existence of MNV by multimodal imaging and differentiate non-neovascular lesions with HRM from ICNV or AMD.

There are also some cases of non-neovascular pachychoroid lesion with HRM, which have SRF, as shown in Fig. 6. We speculate that the pathological mechanism of these cases are similar to our cases without SRF. However, these cases are now generally categorized as CSC due to SRF and photodynamic therapy might be taken into consideration for treatment [33].

The limitations of this study are the small sample size and its retrospective nature. Although these cases are very rare in the clinical setting, larger numbers of study subjects with detailed analysis are needed in the future. In addition, the evaluation of multimodal imaging could be reader dependent; to minimize this confounding factor, an AI machine system that includes machine-learning algorithms should be investigated. This could enable earlier detection of non-neovascular pachychoroid diseases and demonstrate there is no need for treatment.