Background/Aims To analyse the long-term anatomic and visual outcomes of patients with peripapillary pachychoroid syndrome (PPS), a recently described entity in the pachychoroid disease spectrum.
Methods This study retrospectively included patients from several retina centres worldwide. Visual acuity (VA), retinal thickness and choroidal thickness at baseline, 6 months and final follow-up were assessed. Temporal trends in VA and anatomic characteristics were evaluated. Visual and anatomic outcomes in eyes that were observed versus those that were treated were analysed.
Results Fifty-six eyes of 35 patients were included with mean follow-up of 27±17 months. Median VA was 20/36 at baseline and remained stable through follow-up (p=0.77). Retinal thickness significantly decreased subfoveally (p=0.012), 1.5 mm nasal to the fovea (p=0.002) and 3.0 mm nasal to the fovea (p=0.0035) corresponding to areas of increased thickening at baseline. Choroidal thickness significantly decreased subfoveally (p=0.0030) and 1.5 mm nasal to the fovea (p=0.0030). Forty-three eyes were treated with modalities including antivascular endothelial growth factor injection, photodynamic therapy, and others. VA remained stable in treated eyes over follow-up (p=0.67). An isolated peripapillary fluid pocket in the outer nuclear layer was characteristic of PPS.
Conclusion Patients with PPS experienced decreased retinal oedema and decreased choroidal thickening throughout the course of disease. While some patients experienced visual decline, the overall visual outcome was relatively favourable and independent of trends in retinal or choroidal thickening.
Data availability statement
Data are available upon reasonable request.
Statistics from Altmetric.com
The pachychoroid disease spectrum (PDS) is a group of disorders associated with choroidal thickening and hyperpermeability comprised of four major groups: pachychoroid pigment epitheliopathy, central serous chorioretinopathy (CSCR), pachychoroid neovasculopathy and polypoidal choroidal vasculopathy.1–4 The pathophysiological mechanism has been attributed to choroidal vascular hyperpermeability in conjunction with disruption of the retinal pigment epithelium (RPE). These entities can share overlapping features with more than one disorder present in the same eye1 over a wide range of severity.5–7
Peripapillary pachychoroid syndrome (PPS) is a recently reported entity in the PDS spectrum with CSCR-like features that characteristically surround the optic nerve.8 Patients exhibit intraretinal (IRF) and/or subretinal fluid (SRF) emanating from the peripapillary region, chorioretinal folds and choroidal hyperpermeability with pachyvessels. Choroidal thickening is greater in the nasal versus temporal macula in contradistinction to typical CSCR, and eyes with PPS do not necessarily display focal leakage on fluorescein angiography (FA).8 Instead, PPS eyes tend to exhibit mottled fluorescence on FA and fundus autofluorescence.8 The disease shares additional PDS features including hyperopia, serous pigment epithelial detachment, and pigment epitheliopathy.1 3–6 9 10
Herein, we expand on our original paper to describe long-term multimodal imaging, clinical course, visual outcome and treatment response of PPS. We temporally correlate choroidal thickening, retinal fluid accumulation and visual acuity (VA) in these patients to further characterise the condition.
Materials and methods
This retrospective, multicentre cohort study included patients from December 2011 to April 2018.
Inclusion criteria were those fulfilling diagnosis of PPS,8 that is, peripapillary choroidal thickening associated with IRF and SRF in the nasal macula and choroidal hyperpermeability. Eyes with ocular surgery within 6 months, medications associated with macular oedema, vitreomacular traction, epiretinal membrane, diabetic retinopathy, vein occlusion, macular neovascularisation (MNV), cavitary optic disc anomalies or uveitis were excluded. Patients’ age, sex, medical and ocular history, best-corrected VA, phakic spherical equivalent refractive error (if available) and axial length were recorded. We also recorded the type and date of any ocular treatments, including antivascular endothelial growth factor (VEGF) therapy, photodynamic therapy (PDT), focal laser, acetazolamide and others.
Anatomic evolution of the syndrome was primarily characterised by spectral-domain optical coherence tomography (OCT) analysed at three time intervals: at baseline when peripapillary fluid was first identified, at 6 months (±2 months) and at the last available follow-up. Patients with less than 4 months of follow-up were excluded. OCT imaging consisted primarily of horizontal raster or radial scans of the macula, enhanced-depth imaging (EDI, SPECTRALIS, Heidelberg Engineering, Heidelberg, Germany; OCT-HS100, Canon Europe, Uxbridge, UK), and, if available, optic nerve OCT (SPECTRALIS, OCT-HS100, Canon Europe Ltd, Uxbridge, UK).
Additional modalities were reviewed to confirm the diagnosis including fundus photography (Topcon Medical Systems, Oakland, New Jersey; Optos, Dunfermline, Scotland), FA (SPECTRALIS, Topcon, Optos), indocyanine green angiography (ICG) (SPECTRALIS, Optos), and FA (SPECTRALIS, Optos).
OCT anatomic measurements were identical to those performed in the original PPS report.8 Retinal and choroidal thickness measurements were performed at the following positions: subfoveally, 1.5 mm nasal and temporal to the fovea, and 3 mm nasal and temporal to the fovea. In addition, choroidal thickness was measured 250 µm temporal to Bruch’s membrane origin temporal to the disc (BMO250) (figure 1A).
Choroidal thickness was measured from the mid-height of the RPE to the outer boundary of the Haller vessels on the subfoveal OCT scan. Adjustment of image contrast and adjacent OCT cuts were used to help identify the boundary, but if it could not be identified, then the case was excluded. Retinal thickness was calculated in the following manner: the inner boundary of the internal limiting membrane was segmented using a custom Matlab script (MathWorks, Natick, Massachusetts) to calculate the summed height of the retina and SRF from the RPE boundary. All outputs were meticulously reviewed by human graders and manually corrected if needed. Pixel resolution was converted to microns using the image scale bars. OCT images were qualitatively evaluated for the presence of IRF and SRF and presence of a peripapillary fluid pocket (figure 1B).
All quantitative and qualitative analysis was performed by two independent observers (DX and EG), and inter-reader agreement was assessed by interclass correlation (ICC). The averaged measurement between graders was used for analysis. Large quantitative disagreements (>100 µm thickness) and any qualitative disagreements were resolved by open adjudication by the graders and the senior author (DS).
VA was converted to logarithm of the minimum angle of resolution and compared between time points by Wilcoxon signed-rank test. Mean choroidal and retinal thickness were compared using paired t-test. The proportion of patients at each time point with subfoveal and nasal macular SRF or IRF was compared by McNemar’s test. To illustrate clinical differences related to treatment, we divided the cohort into those that were treated by any modality (anti-VEGF, PDT, or others) versus those that were observed and compared VA and anatomic differences between the groups. Statistical analysis was performed in R V.3.6.0. A p value less than 0.05 was considered significant.
A total of 66 eyes of 40 patients were initially eligible for inclusion and displayed characteristic findings of PPS including peripapillary choroidal thickening and IRF or SRF extending from the optic disc. Among these, 4 eyes were excluded because of short follow-up duration, 5 eyes because of poor image quality and 1 eye because of MNV leaving 56 eyes of 35 patients for final analysis. Among these, 14 (40%) cases were unilateral and 21 (60%) cases were bilateral. OCT imaging was analysed at baseline, 6 months (available in 48 of 56 eyes (86%)) and last available follow-up (mean 27±17 months). Patients’ demographic factors are outlined in table 1.
The median VA was 20/36 (IQR 20/25–20/63) at baseline, 20/32 (IQR 20/24–20/53) at 6 months, and 20/32 (IQR 20/25–20/53) at last follow-up. VA remained stable at 6 months (p=0.10) and last follow-up (p=0.77) compared with baseline. At baseline, one eye (2%) exhibited VA worse than 20/200 (due to IRF and ellipsoid zone (EZ) disruption), no eyes were worse than 20/200 at 6 months, and 1 eye (2%, different patient) was worse than 20/200 at last follow-up (due to central RPE atrophy). Figure 2 highlights the typical VA trend for the entire PPS cohort.
We analysed choroidal and retinal thickness at each time point (table 2). The ICC of the graders was excellent at 0.94. Only 1.9% of measurements displayed significant discrepancy requiring adjudication. Retinal thickness significantly decreased in the first 6 months at every measured location, while at last follow-up there was a significant reduction in retinal thickness in the subfoveal, nasal 1.5 mm and nasal 3.0 mm locations corresponding to where oedema is typically worst in PPS (table 3). There was a significant reduction in the subfoveal and nasal 1.5 mm choroidal thickness at last follow-up. Baseline subfoveal retinal (p=0.40) and choroidal thickness (p=0.96) had no statistical association with VA. Subfoveal retinal thickness (SRT) decreased in 34 of 48 eyes (71%) over the first 6 months and 36 of 56 eyes (64%) over follow-up. Eyes with decreased SRT at 6 months were more likely to maintain that decrease at last follow-up (p=0.016). At 6 months, eyes with decreased SRT had improved vision compared with those without by a mean of 5.5±11 letters (p=0.023), although there was no difference at last follow-up (p=0.27).
We additionally analysed the proportion of patients with SRF or IRF subfoveally and in the nasal macula (table 4). There was a significant reduction in subfoveal SRF (p=0.0098), nasal macula SRF (p=0.046) and nasal macula IRF (p=0.0098) at 6 months. This reduction was maintained at last follow-up in the same locations (subfoveal SRF, p<0.001; nasal macula SRF, p=0.043 and nasal macula IRF, p=0.015). VA was significantly worse in patients with subfoveal SRF at baseline (mean 20/50 vs 20/33, p=0.019) and a trend towards worsening with subfoveal IRF (mean 20/65 vs 20/36, p=0.054).
Post-hoc review of the OCT images revealed a commonly occurring small cystoid space immediately temporal to the optic disc margin which we termed a peripapillary fluid pocket (PFP, figure 1B) that may we believe may be a biomarker associated with PPS. This isolated fluid pocket displayed a characteristically uniform size and was located within the outer nuclear layer (ONL) and was separated from the diffuse IRF located nasal and subfoveal in the macula. Presence of this characteristic fluid pocket could be identified in 25 eyes (44%) in our cohort and was significantly associated with nasal macular IRF (OR=5.1, p<0.001) and peripapillary atrophy (OR=4.7, p=0.0083). We did not find an association with SRF or choroidal thickening.
Of the 56 eyes, 13 eyes (23%) were observed without treatment (example shown in figure 3). Baseline and final VA were similar at 20/30 (IQR 20/20–20/60) and 20/38 (IQR 20/25–20/50, p=0.44), respectively. The remaining 43 eyes (77%) received variegated treatment modalities including anti-VEGF injections (figure 4), photodynamic laser therapy (PDT, figure 5), focal macular laser, oral or topical carbonic anhydrate inhibitors, or topical or intravitreal steroid. The median baseline VA (20/40, IQR 20/30–20/63) was similar to final VA (20/32, IQR 20/25–20/56, p=0.67).
The most common treatments were anti-VEGF therapy and PDT. Twenty-three eyes (41%) received a mean of 4.6±3.4 anti-VEGF injections (bevacizumab, ranibizumab or aflibercept) over 26 months. SCT decreased by 50±58 µm (p<0.001) while SRT decreased by 88±160 µm (p=0.015) at final follow-up, although VA was stable (p=0.57). Eight eyes (14%) received PDT. There was a trend toward decreased SRT in PDT eyes (−144±189 µm, p=0.070) and no change in SCT (p=0.66), while VA remained stable (p=0.42).
PPS is a newly reported entity in the PDS. Our initial study8 described the defining features including peripapillary IRF and SRF associated with significant choroidal thickening in the nasal macula. Additional characteristics include choroidal folds, peripapillary RPE mottling and pachyvessels on ICG. The peripapillary predilection associated with nasal choroidal thickening, a crowded disc appearance, and peripapillary IRF distinguish this disorder from typical CSCR. These findings were again identified in the current study’s PPS cohort recruited from a number of retinal centres worldwide.
Our analysis demonstrated an overall significant decrease in retinal and choroidal thickness over time in the nasal macula which colocalised with the baseline area of greatest retinal thickness. Despite the reduction in retinal oedema, there was no overall vision gain, which may have been due to several factors. While subfoveal SRF significantly decreased, subfoveal IRF persisted without change, and overall 25% of eyes retained either subfoveal IRF or SRF. More importantly, some PPS eyes developed subfoveal photoreceptor or RPE atrophy leading to visual decline despite resolution of fluid. In the eight eyes with acuity worse than 20/100 at the final visit, 6 (75%) displayed either EZ attenuation or RPE atrophy. This outcome was similar to typical CSCR patients in whom a portion develop severe outer retinal and RPE disruption.11 12
We compared patients who were observed vs those that received treatment with anti-VEGF or PDT. Overall, there was no visual benefit in the treated group. Subfoveal retinal and choroidal thickness decreased after anti-VEGF therapy at final follow-up. However, like CSCR, PPS cases may exhibit a self-limiting course without treatment while others have persistent fluid despite treatment. Studies of anti-VEGF therapy versus observation in CSR have yielded mixed results regarding central macular thickness and generally found no statistical difference in visual outcome which is in agreement with the observations in this paper.13–17 However, a reduction in choroidal thickness with anti-VEGF, PDT and observation have been noted in other CSCR studies, similar to our observations with PPS.18–21
Typical CSCR is mediated by increased choroidal hydrostatic pressure, permeable choroidal vessels and a break in the RPE leading to subretinal fluid accumulation.10 While SRF is a sine qua non-feature of acute CSCR, the later development of IRF may be the result of secondary factors including inflammatory, degenerative or neovascular complications.22–24 The temporal trend of IRF and SRF in PPS may shed light on its pathogenesis. IRF was more prevalent than SRF and appeared earlier in the disease course in contrast to typical CSCR. While choroidal hydrostatic pressure is likely a factor in PPS, the presence of peripapillary atrophy and unique peripapillary circulation may also play a role in the pathophysiology. The absence of RPE around the nerve may lead to direct transmission of choroidal hydrostatic pressure on the retina explaining the greater association of IRF with PPS. Also, arteriolar branches from the choroid form the circle of Zinn-Haller, a cilioretinal anastomotic vascular complex that perfuses the prelaminar optic nerve fibres.25 26 This anastomosis may transmit choroidal hydrostatic pressure directly to the inner retina in the peripapillary region. Thus, we believe that RPE atrophy and the peripapillary arteriolar anastomosis may explain the unique peripapillary location of PPS.
We additionally observed a peripapillary collection of cystic fluid, termed the peripapillary fluid pocket, in about half of the eyes. This isolated pocket was always noted in a characteristic location within the ONL immediately temporal to the disc and was distinct from the diffuse IRF in the outer plexiform layer. Presence of a PFP correlated with worse IRF and was significantly associated with peripapillary atrophy. This finding has not been described in CSCR or other pachychoroid spectrum disorders. We hypothesise that the PFP may represent the entry site of fluid from the choroid to the retina and may be a biomarker of PPS related to the mechanism of peripapillary fluid accumulation.
There are several limitations of this study given its retrospective nature, heterogeneity of treatments and relatively small sample size. Patients may have received more aggressive treatment with more severe disease, limiting or masking any treatment benefit. Further, treatment benefit may not have been detected due to a ceiling effect associated with excellent baseline VA. Choroidal thickness measurements in eyes with extremely thick choroid were limited by poor contrast and obscuration of the choroidal-scleral boundary, however these images were excluded from the study. A single subfoveal horizontal line scan was used for this analysis and therefore PPS features in extrafoveal locations may have been missed. Nonetheless, the foveal and peripapillary anatomy were both adequately captured and this was correlated temporally in tracked scans. Future studies using swept source volumetric OCT scans may yield more detailed choroidal maps in PPS.
In summary, we described the long-term outcomes of patients with the novel syndrome PPS. Overall, these eyes displayed a decrease in subfoveal SRF and nasal macular IRF and SRF over time. Choroidal thickness also decreased especially in the nasal macula. While some patients did experience significant vision decline, overall the natural history and VA outcomes of this new syndrome were favourable and observation is a viable management option. Treatments such as PDT and anti-VEGF therapy may be considered in select patients, especially those with subfoveal subretinal fluid, and were associated with improved retinal and choroidal anatomy but overall VA was unchanged. Larger prospective evaluations regarding treatment are necessary to make definitive conclusions.
Data availability statement
Data are available upon reasonable request.
The study was approved by the Institutional Review Board (IRB) of the University of California Los Angeles and adhered to the tenets of the Declaration of Helsinki. IRB approval was obtained by the coauthors at their respective institutions. The study complied with the Health Insurance Portability and Accountability Act.
Correction notice This article has been corrected since it first published. The provenance and peer review statement has been included.
Contributors The research design was created by DX, EG, NP, WKL, and DS. All members participated in data collection, data analysis was performed by DX, EG, and DS. All members participated in drafting and critical revisions of the manuscript.
Funding SS: Heidelberg Engineering, Carl Zeiss Meditec, Optos, Centervue, Novartis, Bayer, Boeheringer, Allergan, Alcon, Topcon, Nidek. KBF and DS: Macula Foundation Inc, New York NY.
Competing interests NP: speaker for Allergan, Bayer, Novartis and Optovue. SL: consultant for Novartis, personal fees and non-financial support from Heidelberg Engineering and Carl Zeiss Meditec, personal fees from Novartis, Allergan and Bayer. WKL: consultant for Bayer Healthcare, Novartis, and Santen Pharmaceutical. SS: consultant for Heidelberg Engineering, Zeiss Meditec, Optovue, Optos, Centervue, Nidek, Novartis, Bayer. KBF: consultant for Zeiss, Optovue, Novartis, Allergan, and Heidelberg Engineering. DS: Consultant for Amgen, Bayer, Genentech, Heidelberg, Novartis, Optovue, Topcon, Regeneron.
Patient and public involvement statement Not commissioned; externally peer reviewed.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.