Pachychoroid Disease Progression: Two-Year Outcomes and OCT Features Associated With Neovascularization

The pachychoroid disease spectrum includes several phenotypes associated with choroidal abnormalities, including uncomplicated pachychoroid (UP), pachychoroid pigment epitheliopathy (PPE), central serous chorioretinopathy (CSCR), and pachychoroid neovasculopathy (PNV). Although increased choroidal thickness and dilated Haller-layer vessels are characteristic features, choroidal thickness alone does not reliably distinguish healthy from diseased eyes, and no universally accepted threshold defines abnormal thickness.

The investigators sought to quantify diagnostic progression across pachychoroid subtypes over 2 years and identify baseline clinical or multimodal imaging features associated with progression and incident neovascularization. The study extended a previously reported 1-year cohort.

• Design: Prospective, multicenter, longitudinal observational cohort with 104 weeks of follow-up.
• Setting: Three tertiary eye hospitals: Singapore National Eye Centre, National University Hospital, and Hong Kong Eye Hospital.
• Recruitment period: July 2020 to November 2023.
• Eligibility: Participants aged 45 years or older with SFCT of at least 300 μm measured beneath the foveal center.
• Analysis cohort: 205 eyes from 125 participants completed the 2-year follow-up; mean age was 59.0 ± 8.8 years and 32 participants (25.6%) were female.
• Baseline diagnoses: UP, 39 eyes (19.0%); PPE, 102 (49.8%); CSCR, 47 (22.9%); and PNV, 17 (8.3%). Eyes with polypoidal choroidal vasculopathy at baseline were excluded.

Of the 239 eyes enrolled, 223 completed year 1 and 205 completed year 2. The principal analysis included those completing the second year, so attrition and the selected eligibility criteria should be considered when interpreting generalizability.

All eyes underwent standardized clinical assessment and multimodal imaging. The protocol included macular spectral-domain optical coherence tomography (SD-OCT; 6 × 6 mm cube with 91 B-scans), OCT angiography (OCTA; 6 × 6 mm, 512 × 512 resolution), optical biometry, fundus autofluorescence (FAF), and color fundus photography. Imaging modalities were graded independently by two readers, with disagreements adjudicated by another study author.

• UP: Increased choroidal thickness without RPE disturbance or retinal fluid.
• PPE: RPE disturbance on OCT or FAF without retinal fluid.
• CSCR: RPE alterations with evidence of active or resolved subretinal fluid, intraretinal fluid, or pigment epithelial detachment.
• PNV: Neovascularization detected on OCTA or indocyanine green angiography, with or without fluid.

The primary outcome was a change from baseline pachychoroid subtype to another category at any follow-up visit over 24 months. Secondary outcomes included incident neovascularization, changes in BCVA, SFCT and central subfield thickness (CST), and treatment initiation. Assessments were scheduled at baseline and months 6, 12, 18, and 24.

• Overall progression: 15 of 205 eyes (7.3%; 95% CI, 4.5%–11.6%) changed diagnostic category during 2 years.
• Timing: Twelve eyes (5.9%) had already progressed by year 1. Only three additional eyes developed a diagnostic change between year 1 and year 2.
• Most eyes remained stable: 190 of 205 eyes (92.7%) had no change in diagnosis at year 2 compared with baseline.
• Progression from UP: No UP eye progressed to CSCR or PNV. One of 39 UP eyes (2.6%) progressed to PPE, during the first year.
• Progression from PPE: The additional year-2 changes included two eyes progressing from PPE to CSCR and one from PPE to PNV. One eye that had progressed from PPE to CSCR in year 1 subsequently progressed to PNV in year 2.

The pattern indicates that most observed diagnostic transitions occurred during the first year. The small number of later events limits conclusions about the timing of progression beyond this cohort and follow-up period.

Incident neovascularization was detected in 7 eyes (3.4%; 95% CI, 1.7%–6.9%). Every eye developing neovascularization had shallow, irregular RPE elevation at baseline: 7 of 116 eyes with this feature (6.0%) developed PNV, compared with 0 of 89 eyes without it (0%; 95% CI for the latter, 0%–4.1%). This is a notable association, although the small number of events means the estimate remains imprecise.

In multivariable analysis, female sex was associated with higher observed rates of diagnostic progression (13.5% vs 5.2% in males; P = .02) and incident neovascularization (7.7% vs 2.0%; P = .02). The sex risk ratios in the table are presented using the opposite reference direction to the narrative comparison; the percentages and reported P values are therefore stated here without reversing the tabulated ratios. These associations should be interpreted cautiously because there were few events, confidence intervals were wide, and the cohort was predominantly male. Shallow, irregular RPE elevation showed an adjusted RR of 3.43 for diagnostic progression (95% CI, 0.89–13.30; P = .07), which did not reach conventional statistical significance.

• Anti-VEGF treatment: Three of the seven eyes developing PNV (42.9%) received anti-VEGF injections. The reported regimens included bevacizumab, ranibizumab, and aflibercept.
• Observation: Four of seven PNV eyes (57.1%) were managed without anti-VEGF treatment.
• Photodynamic therapy: Two additional eyes that progressed from PPE to CSCR or PNV required half-fluence photodynamic therapy.

These treatment observations describe management in this cohort; they do not establish a universal treatment threshold for PNV.

• BCVA: Mean BCVA changed from 0.13 ± 0.18 logMAR at baseline to 0.11 ± 0.24 logMAR at year 2, an estimated difference of −0.02 logMAR (approximately one letter; 95% CI, −0.05 to 0.02; P = .33). This was not statistically significant.
• SFCT: Mean SFCT changed from 419 ± 83 μm to 424 ± 90 μm; difference, 6 μm (95% CI, −5 to 16; P = .32). There was no significant overall change.
• CST: Mean CST decreased from 248 ± 41 μm to 238 ± 44 μm; difference, −10 μm (95% CI, −17 to −4; P = .002).
• Baseline visual acuity by subtype: Eyes with CSCR and PNV had worse baseline BCVA than eyes with UP and PPE. Median baseline BCVA was approximately 20/25 for UP and PPE and 20/32 for CSCR and PNV.

Overall, the cohort showed preserved mean visual acuity and no significant increase in choroidal thickness over 2 years. The statistically significant reduction in CST should be distinguished from the nonsignificant BCVA and SFCT changes.

• Review RPE morphology carefully: Shallow, irregular RPE elevation was present at baseline in all eyes that subsequently developed neovascularization. Its presence may support closer surveillance and a lower threshold for multimodal imaging, including OCTA, particularly in eyes with PPE.
• Do not equate a thick choroid with inevitable progression: Most eyes remained in the same diagnostic category over 2 years, and baseline SFCT did not differ significantly among diagnostic groups.
• Interpret OCTA in clinical context: The study used OCTA and indocyanine green angiography to identify neovascularization. A suspicious RPE configuration merits careful assessment rather than assuming that every pachychoroid eye has active neovascular disease.
• Individualize follow-up: The authors suggest that eyes without shallow, irregular RPE elevation may not require prolonged intensive monitoring, if this observation is confirmed in larger studies. Eyes with this feature, especially those with PPE, may warrant closer surveillance.
• Treatment should reflect activity and clinical findings: More than half of the eyes with incident PNV were observed without anti-VEGF treatment in this cohort. The study does not define a treatment algorithm; treatment decisions remain individualized.

Clinical caution: The authors present the RPE feature as a potentially useful risk-stratification marker, not as a validated stand-alone rule for deciding follow-up intervals or treatment.

Strengths

• Prospective, multicenter, 2-year follow-up across three tertiary ophthalmic centers.
• Standardized serial clinical assessments and multimodal imaging.
• Independent grading of imaging modalities by two readers, with adjudication of disagreements.
• Statistical methods accounted for correlation between two eyes from the same participant.

Limitations

• The number of diagnostic progression and neovascularization events was small, limiting statistical power and producing wide confidence intervals.
• Progression occurring after the 2-year observation period cannot be excluded.
• Participants were at least 45 years old; findings may not generalize to younger patients, including younger individuals with CSCR.
• The SFCT threshold of 300 μm was an operational inclusion criterion, not a universally accepted definition of pathological choroidal thickening. SFCT can vary with age, axial length, and time of day.
• BCVA was measured with ETDRS charts at one center and Snellen charts at the other two, potentially introducing measurement variability.
• Only participants completing year 2 were included in the principal analysis.

Conclusion: In this relatively small prospective cohort, pachychoroid eyes were generally stable over 2 years: 92.7% showed no diagnostic change, and mean BCVA and SFCT did not significantly worsen. Neovascularization was uncommon and occurred only in eyes with shallow, irregular RPE elevation at baseline. This OCT feature may help refine risk stratification and identify eyes that merit closer surveillance, but confirmation in larger cohorts is needed before changing follow-up strategies.