Ocular manifestations in newly diagnosed acute leukemia: diagnostic performance of age, hemoglobin and platelets for retinal haemorrhages.

Abstract

Background: Ocular involvement in the diagnosis of leukemia is clinically important but has been reported variably. Our aim was to define its prevalence and spectrum and to identify hematologic predictors, specifically age, hemoglobin level (Hb), platelet count (PLT), mean platelet volume (MPV), and platelet mass index (PMI).

Methods: This retrospective, single-center study evaluated all newly diagnosed leukemia patients between January 2023 and March 2025. All patients were examined by an ophthalmologist at the time of diagnosis and, if present, ocular involvement was identified.

Results: Among 78 pediatric patients, ocular involvement was present in 17 of 78 patients (21.8%). The most common ocular manifestation was retinal hemorrhage (16.7%). Ocular involvement was more frequent in acute myeloid leukemia (AML) than in acute lymphoblastic leukemia (ALL) (46.7% vs. 15.9%, p= 0.016). Older age (for ocular involvement: median 13 years [6-15] vs. 5 years [3-11], p = 0.006; for retinal hemorrhage: median 14 years [10-15] vs. 5 years [3-11], p = 0.001) and lower hemoglobin (for ocular involvement: 6.68 ± 2.31 g/dL vs. 8.46 ± 2.47 g/dL, p = 0.01; for retinal hemorrhage: 6.33 ± 2.26 g/dL vs. 8.43 ± 2.46 g/dL, p= 0.006) were associated with a higher risk. In multivariate models, age and Hb were independent predictors of hemorrhage (age: odds ratio [OR] 1.29, 95% confidence interval [CI] 1.10-1.52 per year; Hb: OR 0.60, 95% CI 0.38-0.95 per g/dL). ROC analyses showed that age provided the best discrimination (area under the curve [AUC] 0.796; 95% CI 0.651-0.940; cut-off ≥10 years: 73% sensitivity, 70% specificity), followed by Hb (AUC 0.756; 95% CI 0.604-0.908; cut-off 6.95 g/dL: sensitivity 76%, specificity 76%), whereas PLT showed weaker discrimination (AUC 0.682; 95% CI 0.553-0.812; cut-off 48,000/mm³: sensitivity 62%, specificity 65%).

Conclusion: Approximately one in five children had ocular involvement at diagnosis, most commonly retinal hemorrhage. Age and hemoglobin independently predicted retinal hemorrhage and provided useful discrimination. These data support a "hemoglobin-first" triage that prioritizes initial and follow-up ocular fundus testing in older children and those with an Hb ≲7 g/dL, with external validation of the proposed thresholds warranted.