Diagnostic performance of conventional optical coherence tomography parameters for differentiating pediatric papilledema from pseudopapilledema

Abstract

Abstract Purpose To evaluate the diagnostic performance of conventional optical coherence tomography (OCT)-derived peripapillary retinal nerve fiber layer (RNFL) and macular ganglion cell-inner plexiform layer (GC-IPL) thickness measurements for differentiating papilledema from pseudopapilledema in children. Methods This retrospective study included 115 pediatric patients (230 eyes) with papilledema, optic disc drusen (ODD)-related pseudopapilledema, pseudopapilledema without optic disc drusen, and healthy controls. Peripapillary RNFL and macular GC-IPL thicknesses were measured using spectral-domain OCT. Group comparisons were performed using generalized estimating equations, and diagnostic performance was assessed using receiver operating characteristic analysis. Results The study included 26 patients (52 eyes) with papilledema, 30 patients (60 eyes) with optic disc drusen (ODD)-related pseudopapilledema, 32 patients (64 eyes) with pseudopapilledema without ODD, and 27 healthy controls (HCs) (54 eyes). Mean pRNFL thickness was 129.85 ± 51.08 μm in the papilledema group, 128.22 ± 36.78 μm in the ODD-related pseudopapilledema group, 106.64 ± 17.10 μm in the pseudopapilledema without ODD group, and 99.89 ± 11.29 μm in the healthy control group. Compared with HCs, RNFL thickness was significantly increased in all quadrants in eyes with ODD-related pseudopapilledema, in three quadrants in papilledema, and only in the superior quadrant in pseudopapilledema without ODD. Compared with HCs, inferior RNFL demonstrated the highest discriminatory ability for papilledema (AUC = 0.848, p < 0.001) and ODD-related pseudopapilledema (AUC = 0.906, p < 0.001), whereas superior RNFL showed significant discrimination for pseudopapilledema without ODD (AUC = 0.720, p = 0.001). For differentiating papilledema from ODD-related pseudopapilledema, nasal RNFL yielded the highest AUC among the evaluated RNFL parameters (AUC = 0.699, p = 0.005). For papilledema versus pseudopapilledema without ODD, inferior RNFL demonstrated the highest diagnostic performance (AUC = 0.732, p < 0.001), followed by temporal RNFL (AUC = 0.710, p = 0.002). Several GC-IPL parameters demonstrated significant discriminatory ability in specific comparisons, with inferonasal GC-IPL showing the highest AUC for ODD-related pseudopapilledema versus healthy controls (AUC = 0.731, p = 0.004) and superotemporal GC-IPL showing the highest AUC for papilledema versus ODD-related pseudopapilledema (AUC = 0.769, p = 0.001). Conclusions Conventional OCT-derived RNFL measurements demonstrated limited diagnostic performance for distinguishing pediatric papilledema from pseudopapilledema, particularly in eyes with ODD-related pseudopapilledema. OCT findings should therefore be interpreted together with clinical assessment and, where available, advanced optic nerve head imaging techniques.