SHRM in Central Serous Chorioretinopathy: Persistent Fluid, Photoreceptor Damage, and Visual Outcomes

Study design: Retrospective, multicenter comparative cohort assembled through the Macula Society International CSCR Research Network (MICRoN); records collected from January 2010 through December 2023.

Sample: 503 eyes of 503 patients: 103 with SHRM and 400 without SHRM. Mean age was 47.1 ± 11.2 years, and median symptom duration was 2 months (IQR 0.5–6). Median follow-up was 20 months (IQR 8–56).

Primary questions: How does OCT-detected SHRM relate to longitudinal best-recorded visual acuity (BRVA), persistent subretinal fluid (SRF), macular neovascularization (MNV), and post-resolution photoreceptor integrity?

Important eligibility detail: Eyes with MNV at baseline were excluded, but new or concurrent MNV detected during the study was analyzed.

SHRM denotes medium- to high-reflectivity material between the neurosensory retina and retinal pigment epithelium (RPE) on OCT. Its composition in CSCR remains uncertain: fibrinous exudation, photoreceptor-derived debris, or mixed mechanisms are proposed but not established histologically.

• Grade 1: SHRM occupies less than 20% of the subretinal space (38 eyes).
• Grade 2: SHRM occupies 20%–50% (55 eyes).
• Grade 3: SHRM occupies more than 50%, sometimes with layered or onion-like reflectivity (10 eyes).

Investigators manually delineated SHRM in selected horizontal OCT scans, including the fovea and areas of maximal accumulation. The three grades quantify area occupancy, not material composition or a validated treatment-response score.

Eyes with SHRM had worse baseline BRVA than eyes without SHRM: 0.4 ± 0.3 vs 0.2 ± 0.3 logMAR (approximately 20/50 vs 20/30; Table 1, P = .001). The abstract reports P = .006 for this comparison; the main results and Table 1 report P = .001.

Compared with non-SHRM eyes, SHRM eyes had greater RPE alteration area (3.2 vs 1.9 disc areas; P = .04), higher neurosensory retinal detachment (NSRD) height (270.7 vs 123.6 µm; P < .001), more photoreceptor outer-segment (PROS) irregularity (46.6% vs 28.8%; P = .004), more hyperreflective foci in PROS (99% vs 59.5%; P < .001), and more double-layer sign (56.3% vs 32%; P = .001). Subfoveal choroidal thickness did not differ significantly (P = .23).

SHRM should therefore be interpreted in the context of the full OCT and multimodal imaging phenotype, rather than as an isolated diagnosis.

The study classified disease course as resolution, recurrence, or persistence. In SHRM eyes, 36/103 (34.9%) resolved, 25/103 (24.3%) recurred, and 42/103 (40.8%) remained persistent. In non-SHRM eyes, the corresponding proportions were 206/400 (51.5%), 124/400 (31.0%), and 70/400 (17.5%). The overall distribution differed (P < .001), driven in particular by greater persistence with SHRM.

At final follow-up, residual SRF was observed in 47/103 SHRM eyes (45.6%) versus 106/400 non-SHRM eyes (26.5%; P < .001). This final-visit SRF measure is distinct from the study’s three-category disease-course classification and should not be conflated with the 40.8% vs 17.5% persistence rates.

In adjusted logistic regression, non-SHRM status was associated with lower odds of persistent SRF relative to SHRM (OR 0.36, 95% CI 0.19–0.77; P = .001). Baseline RPE alteration area had a borderline association (OR 1.13, 95% CI 1.00–1.29; P = .05). These observational associations do not demonstrate that SHRM causes persistence.

Both cohorts improved significantly in BRVA from baseline to the final visit (within-group P < .001). Final mean BRVA was approximately 0.2 ± 0.3 logMAR in each group, without a significant between-group difference (P = .67), despite worse initial acuity in the SHRM cohort.

Within the SHRM cohort, eyes with residual SHRM had worse final vision than those in which SHRM was absent at follow-up (reported 0.4 ± 0.4 vs 0.1 ± 0.3 logMAR; P = .02), although the comparison is observational and may reflect underlying disease severity.

Linear regression associated age, symptom duration, baseline BRVA, SHRM-group status, and persistent or recurrent disease course with change in BRVA. The regression coefficient signs depend on the study’s definition of visual-acuity change; they should not be interpreted as causal effects of treatment or as a reliable individualized prediction formula.

Among eyes assessed after complete SRF resolution, ellipsoid-zone (EZ) loss occurred in 25/56 SHRM eyes (44.6%) versus 82/294 non-SHRM eyes (27.8%; P = .01). EZ disruption was defined as at least 100 µm of discontinuity corresponding to areas of previous NSRD.

Among the three SHRM grades, EZ loss was more frequent in Grade 3 than Grades 1 or 2 (overall P = .002; pairwise adjusted P = .009 for both comparisons). Only 10 eyes had Grade 3 at baseline, limiting precision. Eight SHRM eyes (7.8%) developed residual subretinal scarring, including two Grade-3 eyes.

Clinical distinction: Fluid clearance does not guarantee restoration of photoreceptor architecture. Post-resolution EZ integrity should be assessed alongside BCVA.

MNV was detected during follow-up in 25/103 SHRM eyes (24.3%) versus 50/400 non-SHRM eyes (12.5%; P = .01). Baseline MNV was an exclusion criterion. Detection used ICGA or OCT angiography according to local availability.

SHRM is not synonymous with MNV, and its mere presence should not trigger an anti-VEGF treatment decision. However, because neovascular complexes may coexist with or become apparent during CSCR follow-up, suspicious OCT findings warrant appropriate multimodal assessment. The authors note that SHRM itself can attenuate OCTA signals, potentially obscuring MNV.

Grades 1, 2, and 3 comprised 38, 55, and 10 eyes, respectively. Baseline NSRD height was highest in Grade 1 rather than Grade 3, showing that SHRM occupancy and fluid height are not interchangeable markers. Persistent SRF did not differ significantly across grades (39.4%, 43.6%, and 30%, respectively).

Four Grade-1 eyes progressed to Grade 2, and six Grade-2 eyes progressed to Grade 3 at follow-up. These changes occurred alongside varying SRF behavior, suggesting possible reorganization of the subretinal material rather than a simple linear measure of active leakage. The small numbers and inconsistent follow-up schedules prevent reliable progression-rate estimates.

At presentation: Document SHRM presence, extent, and associated PROS abnormalities, RPE change, NSRD height, and possible MNV. Grade the deposit only if OCT quality and scan coverage are adequate.

At follow-up: Track both SRF and SHRM, reassess EZ integrity after fluid resolution, and consider multimodal imaging when MNV is suspected.

For counseling: Explain that SHRM identifies a group with more persistent fluid and a higher burden of structural damage, but meaningful visual recovery remains possible. This study cannot establish that earlier PDT, laser, anti-VEGF, or combination therapy is superior for SHRM-positive CSCR.

• Retrospective, nonrandomized treatment allocation: Observation, PDT, laser, anti-VEGF, mineralocorticoid receptor antagonists, and combinations were used in differing proportions. Confounding by indication precludes comparative treatment-effect conclusions.
• Variable follow-up: Median 20 months with wide dispersion (IQR 8–56); exact SRF resolution and recurrence timing could not be established consistently.
• Imaging heterogeneity: Multiple OCT devices and inconsistent OCTA/ICGA availability may affect measurement and MNV ascertainment.
• Manual SHRM grading: Delineation is observer-dependent, and the paper does not report an independent intergrader reproducibility estimate for this cohort.
• Small Grade-3 subgroup: Ten eyes limit subgroup inference; grade comparisons may be underpowered.
• Potential reporting inconsistencies: The abstract and main results provide different P values for baseline BRVA (P = .006 vs .001). The Results text also reports 36 patients with previous treatment while listing 26 and 67 patients in the two groups; those figures are internally inconsistent, so the aggregate previous-treatment count should not be treated as reliable.
• Interpretive limit: SHRM may be a marker of severity rather than a mechanistic driver of persistent fluid or photoreceptor injury. Prospective validation is needed before assigning treatment thresholds.

MICRoN Report 6 strengthens the evidence that SHRM on OCT identifies a clinically distinct CSCR phenotype: worse presenting vision, more persistent SRF, more frequent incident or concurrent MNV detection, and greater post-resolution EZ loss. Yet mean vision improved in both SHRM and non-SHRM groups, underscoring that SHRM is not a deterministic marker of poor final acuity. The next research step is prospective, standardized imaging with controlled treatment allocation to determine whether SHRM can guide therapy rather than merely stratify prognosis.