A systematic review and meta-analysis of diagnostic test accuracy studies has determined that shallow irregular retinal pigment epithelium elevation phenotypes detected on structural optical coherence tomography demonstrate high summary sensitivity alongside limited specificity for identifying concurrent macular neovascularization in eyes affected by central serous chorioretinopathy. Across nine primary cohorts evaluating 671 eyes, the harmonized structural phenotype achieved a model-based summary sensitivity of 98 percent with a 95 percent confidence interval spanning 77 to 100 percent, whereas the summary specificity reached only 69 percent with a 95 percent confidence interval between 53 and 81 percent. These findings indicate that while shallow irregular elevations of the retinal pigment epithelium serve as sensitive structural indicators of neovascular involvement, their presence does not constitute definitive diagnostic proof of abnormal vessel proliferation, and their absence cannot be reliably interpreted as an absolute rule-out test in clinical practice.
Central serous chorioretinopathy is recognized as a pachychoroid-spectrum disorder marked by serous neurosensory retinal detachment, retinal pigment epithelium dysfunction, and underlying choroidal vascular abnormalities. Although acute episodes frequently resolve without medical intervention, chronic or recurrent disease states can cause persistent subretinal fluid accumulation, extensive pigmentary disturbance, outer retinal disruption, and permanent loss of visual acuity. The emergence of macular neovascularization represents a particularly severe complication of long-standing disease that fundamentally alters clinical management. Patients with neovascular complications may benefit directly from intravitreal antivascular endothelial growth factor administration, whereas non-neovascular manifestations are managed through watchful waiting, systemic risk-factor modification, photodynamic therapy, or targeted laser interventions. Consequently, establishing the presence or absence of abnormal vascular complexes is paramount for determining appropriate therapeutic regimens.
Differentiating neovascular lesions from chronic pachychoroid-related tissue alterations presents persistent diagnostic hurdles for clinicians. Unlike typical neovascular age-related macular degeneration, neovascularization occurring in central serous chorioretinopathy predominantly manifests as type 1 tissue situated beneath the retinal pigment epithelium, where it often remains quiescent or minimally exudative. Traditional diagnostic modalities, including dye-based fluorescein angiography and indocyanine green angiography, frequently encounter interpretative limitations in chronic presentations because widespread pigment epithelial disturbance, transmission window defects, vascular staining, choroidal hyperpermeability, and longstanding fluid accumulations can obscure or mimic active neovascular leakage. Optical coherence tomography angiography offers depth-resolved noninvasive visualization of pathological blood flow signals, yet the technology is not universally available across clinical centers, remains vulnerable to motion and segmentation artifacts, and is often deployed selectively only after structural abnormalities are noted.
To guide clinical triage, ophthalmologists have historically relied on cross-sectional structural imaging to identify morphological warning signs. Researchers have documented several structural patterns suggestive of underlying neovascular networks, designated across clinical literature as flat irregular pigment epithelial detachments, the double-layer sign, irregular pigment epithelial detachments, shallow irregular detachments, and slight elevations presenting with delicate surface undulations. Despite varying terminology, these descriptions share an essential anatomical configuration characterized by a shallow, broad-based, irregular separation of the retinal pigment epithelium from Bruch’s membrane, often housing hyperreflective or heterogeneous sub-epithelial material. However, identical structural elevations can arise from chronic, non-neovascular pachychoroid stress, creating an urgent need to formally evaluate whether these morphological features can reliably confirm or exclude neovascular growth.
The meta-analysis, conducted under a prospectively registered protocol on PROSPERO under registration number CRD420261404129, searched nine major bibliographic databases through May 19, 2026, without language restrictions, identifying 234 total records and retrieving 26 full-text reports for detailed assessment. Eligible investigations were required to examine patients diagnosed with central serous chorioretinopathy, evaluate structural shallow irregular retinal pigment epithelium elevation as an index test, employ optical coherence tomography angiography either alone or embedded within a multimodal framework as the reference standard, and furnish extractable contingency data for true-positive, false-positive, false-negative, and true-negative outcomes. Studies lacking non-phenotypic comparator groups or those evaluating dye angiography without structural index tests were excluded from the primary synthesis. Risk of bias and methodological applicability were formally appraised across patient selection, index test, reference standard, and flow domains using the standardized QUADAS-2 instrument.
The primary quantitative synthesis synthesized evidence from nine studies encompassing 671 eyes, derived from cohorts conducted internationally across France, China, Poland, Egypt, the United Kingdom, South Korea, Austria, and a collaborative network between the United States and Israel. Denominators where reported accounted for at least 537 individual patients, with cohorts demonstrating a high male predominance typical of the condition. Within the primary dataset, reference-standard testing identified neovascularization in 138 eyes, while 533 eyes were verified as non-neovascular. The aggregate data comprised 130 true-positive classifications, 181 false-positive observations, 8 false-negative cases, and 352 true-negative determinations. Descriptive analysis indicated that the overall proportion of index-test-positive eyes was 46.3 percent, representing 311 of the 671 evaluated eyes, with individual study proportions spanning from 23.0 to 76.7 percent.
Statistical modeling was executed via MetaDTA using a bivariate random-effects generalized linear mixed model with a binomial likelihood and logit link function to estimate summary performance parameters while accounting for between-study heterogeneity and interdependencies between sensitivity and specificity. The summary model yielded an overall false-positive rate of 31 percent with a 95 percent confidence interval between 19 and 47 percent. Scenario-based sensitivity analyses demonstrated that summary sensitivity remained stable between 96 and 99 percent across alternative inclusion parameters, whereas summary specificity exhibited wider fluctuations ranging between 61 and 73 percent. In leave-one-out sensitivity iterations, pooled sensitivity ranged from 97 to 100 percent, and specificity spanned 65 to 73 percent, confirming that while sensitivity was robust across models, the precision of specificity estimates remained sensitive to individual study inclusions.
Crucially, the authors highlighted that the near-perfect sensitivity point estimate must be interpreted with extreme caution due to underlying epidemiological and design limitations. Across the entire primary dataset, only eight false-negative observations were recorded, and seven of the nine included studies reported zero false-negative eyes. This absence of false negatives in most studies contributed to the expansive confidence interval of 77 to 100 percent and reflects potential patient-selection and spectrum biases. Many retrospective investigations utilized enriched or phenotype-stratified cohorts rather than consecutive unselected clinical populations. Furthermore, because structural optical coherence tomography scans routinely precede vascular flow imaging in standard clinical pathways, suspicious morphology often directly prompted subsequent diagnostic testing, introducing verification bias and partial incorporation bias that artificially inflated apparent sensitivity.
The moderate specificity observed in the synthesis underscores that shallow irregular retinal pigment epithelium elevation is not anatomically synonymous with neovascular tissue. In non-neovascular presentations, chronic pachychoroid alterations—characterized by venous congestion, choriocapillaris attenuation, and elevated choroidal pressures—can induce localized pigment epithelial stress and non-vascularized elevation from Bruch’s membrane. Furthermore, the researchers noted that a false-positive classification within the meta-analysis signifies an eye with positive structural morphology lacking detectable flow on optical coherence tomography angiography, rather than an unequivocally avascular tissue complex. Because current imaging technology cannot visualize all slow or deeply shadowed microvascular flow, some reference-negative lesions may harbor occult, low-flow neovascularization that evades detection, meaning these categories reflect operational imaging thresholds rather than definitive histological realities.
Heterogeneity in structural nomenclature further complicated clinical interpretation. Individual investigations utilized disparate terminology, including flat irregular pigment epithelial detachment, double-layer sign, irregular retinal pigment epithelium detachment, and flat hyperreflective detachment. Harmonizing these diverse descriptions under a single unified diagnostic umbrella was necessary for quantitative pooling but means the calculated metrics reflect the broad morphological spectrum rather than any isolated sign. Sensitivity analyses that excluded studies employing broader definitions, such as opaque irregular detachments or flat hyperreflective descriptions, confirmed that definitional breadth primarily drives variations in diagnostic specificity, as broader morphological criteria inevitably capture higher proportions of non-neovascular chronic alterations.
Ultimately, the researchers concluded that shallow irregular retinal pigment epithelium elevation functions effectively as a structural warning phenotype indicating the need for careful clinical assessment and confirmatory vascular imaging, rather than serving as an independent justification for immediate therapy. Clinicians should not initiate antivascular endothelial growth factor injections based on structural cross-sections alone without objective vascular confirmation. Likewise, because the sparse false-negative data and potential verification biases preclude treating the sign as an infallible rule-out metric, high clinical vigilance remains necessary when evaluating older patients, individuals experiencing chronic or recurrent fluid, or eyes failing conventional non-angiogenic interventions. Future diagnostic investigations must implement standardized quantitative morphological metrics, consecutive prospective recruitment, and uniform reference standards applied independently of structural findings.
Subject of Research: Diagnostic accuracy of shallow irregular retinal pigment epithelium elevation phenotypes for detecting macular neovascularization in central serous chorioretinopathy
Article Title: Diagnostic Accuracy of Shallow Irregular Retinal Pigment Epithelium Elevation Phenotypes for Detecting Macular Neovascularization in Central Serous Chorioretinopathy: A Systematic Review and Meta-Analysis
Article References: Sabaner, M. C., Gunes, I., Gulyesil, F. F., Anguita, R., Ferro Desideri, L., Cehofski, L. J., Ramtohul, P., van Dijk, E. H. C., & Subhi, Y. (2026). Diagnostic Accuracy of Shallow Irregular Retinal Pigment Epithelium Elevation Phenotypes for Detecting Macular Neovascularization in Central Serous Chorioretinopathy: A Systematic Review and Meta-Analysis. Ophthalmology and Therapy. https://doi.org/10.1007/s40123-026-01532-9
Image Credits: AI Generated
DOI: 10.1007/s40123-026-01532-9
Keywords: central serous chorioretinopathy, macular neovascularization, retinal pigment epithelium, optical coherence tomography, optical coherence tomography angiography, flat irregular pigment epithelial detachment, double-layer sign, diagnostic test accuracy, systematic review, meta-analysis, pachychoroid spectrum, type 1 neovascularization, fluorescein angiography, indocyanine green angiography, hierarchical DTA model, QUADAS-2, choriocapillaris attenuation, photodynamic therapy
News Source: Ophelia Keating. (November 16, 2026). Retinal Elevation Sign Detects Eye Complications but Lacks Definitive Accuracy. Scienmag.



