Ophthalmic manifestations and treatment outcomes in leukaemia: a comparative analysis of clinical spectrum, visual prognosis and predictors of disease remission across subtypes

Abstract

Aim To conduct an analysis of ophthalmic manifestations in leukaemic patients and evaluate their impact on visual outcomes and disease remission. Methods This retrospective, comparative study included 244 patients (378 eyes) diagnosed with leukaemias between January 2016 and 2024. Data included demography, leukaemia type, ophthalmic manifestations, imaging modalities, treatment modalities (ocular and systemic), treatment outcomes (relapses or remission) and final visual acuity. Univariate and multivariate regression analyses were performed to identify correlations and predictive features associated with visual outcomes. Results The mean age of patients was 34.9±22.1 years (range, 1–88 years). Acute lymphocytic leukaemia was the most common subtype (58%), followed by chronic myeloid leukaemia (CML; 20%). Acute forms manifested earlier in life and were more aggressive. The presence of posterior segment manifestations (64%) was more than anterior segment manifestations (52%). Myeloid leukaemia types were more often associated with haemorrhagic manifestations, whereas lymphoid types were associated with non-haemorrhagic manifestations. Univariate regression analysis showed that the presence of epiretinal membrane (OR 0.110, 95% CI 0.012 to 0.951, p=0.04), submacular haemorrhage (OR 0.091, 95% CI 0.010 to 0.763, p=0.027) and type of leukaemia (acute myeloid leukaemia (AML) (OR 3.195, 95% CI 1.697 to 6.012, p<0.000) or CML (OR 6.39, 95% CI 3.130 to 13.041, p<0.000)) had significant association with poor visual outcome. Multivariate regression analysis revealed AML (OR 2.769, 95% CI 1.392 to 5.507, p=0.004) and CML (OR 7.313, 95% CI 3.326 to 16.074, p<0.000) as independent predictors of poor visual outcome. Conclusion Ophthalmic manifestations are major diagnostic and prognostic clues for leukaemias. Pattern recognition facilitates timely recognition, risk stratification, management and disease remission prediction. All data relevant to the study are included in the article or uploaded as supplementary information.