Leukemia is a life-threatening disease of blood and bone marrow malignant disorders which is caused by an abnormal proliferation of immature leukocytes [1][3]. The data from WHO in 2020 reports 550 Cambodian deaths from leukemia, which accounts for 0.61% of total deaths [2]. Ocular involvement of Leukemia is vital for assessment because ocular symptoms may be the first indication of manifestation, relapse, or early worsening of the condition. Up to 90% of leukemia patients manifests ocular involvements, which commonly results in leukemic retinopathy [3].

In 2015, Jacob Koshy et al., a study conducted in India reported the prevalence of Leukemic ophthalmic lesions were found in 43.8% [4]. Retinal and pre-retinal hemorrhage are commonly found in the posterior segment in Leukemia. Besides, Roth spots, white centered hemorrhages resulting from retinal capillary rupture and extrusion of whole blood, can be found in some cases of leukemic retinopathy. In leukemia, platelet adheres to the damaged endothelium, which initiates a coagulation cascade and forms a white lesion, platelet-fibrin thrombus, in the centre of the hemorrhage. The white centre corresponds to an accumulation of leukemic cells [3]. In 2022, Reza Mirshahi et al., a study conducted in Iran reported low platelet count and low hemoglobin count were found to be associated with intraretinal hemorrhages. In patients with ocular signs, higher mortality rates were observed [5].

Despite these reports, there is very limited published data from Cambodia on the correlation between hematological parameters and retinal manifestations in leukemia. Understanding these associations is essential because retinal findings may serve as surrogate markers of systemic disease severity and prognosis. Identifying hematological parameters of retinopathy could help oncologists and hematologists anticipate ocular complications, initiate timely ophthalmic referral, and improve overall patient outcomes.

Therefore, this study aims to explore the factors associated with the development of retinopathy among acute and chronic leukemia patients at Khmer Soviet Friendship Hospital in 2025. Specifically, it examines the correlation between hematological parameters and the presence of retinopathy, thereby contributing local evidence to guide clinical practice and improve prognostic assessment in Cambodian leukemia patients.

This prospective cross-sectional was conducted at Oncology and Ophthalmology department of Khmer-Soviet Friendship Hospital from 1st February 2025 - 31st July 2025 (6 months). The study adhered to the principle of the Cambodian medical ethics and was approved by the National Ethics Committee for Health Research (NECHR) number 045. Inclusion criteria were: 1) patients of any age with a confirmed diagnosis of acute or chronic leukemia, 2) patients referred consecutively from the onco hematology service during the study period; 3) patients with complete hematological records (hemoglobin, total leukocyte count, platelet count). Patients with a history of diabetes, hypertension, any intraocular surgeries in the past 3 months, autoimmune disorders, vascular occlusions, or hematological disorders other than leukemia were excluded from the study.

Diagnosis was established through complete blood counts and peripheral smear examination; in cases where bone marrow aspiration was not feasible, persistent hematologic abnormalities (marked leukocytosis, anemia, and/or thrombocytopenia) together with compatible clinical features were accepted as diagnostic criteria.

Due to administrative delays in the initial study site, the research was relocated to Khmer-Soviet Friendship Hospital, which also contains oncology and ophthalmology services. As a result, the final sample size was limited to 23 patients.

All information of each patient was collected and recorded in a checklist. All data of the patients were collected under an informed consent form and will be kept confidentially.

Consent form was obtained from all patients who have received the eyes examination, including: name, gender, age, refractive status, history of ocular trauma or surgery, record of hematological parameters including hemoglobin (Hb), total leukocyte count (TLC), platelet count, duration of illness and treatment.

Visual acuity was assessed using the Early Treatment Diabetic Retinopathy Study (ETDRS) chart, and vision categories were defined according to ETDRS scoring standards [6]. In addition, the value of complete blood count was based on the reference intervals provided by the clinical laboratory at Khmer Soviet Friendship Hospital, which are consistent with El Brihi J et al. [7].

The complete ophthalmic examination was done including best-corrected visual acuity (BCVA) on Snellen’s chart, slit lamp biomicroscopy for the anterior segment and posterior segment evaluation by + 90 D (Volk Optical Inc.), and indirect ophthalmoscope (Keeler Ltd., Winsor, UK). Posterior segment examination was performed after pupillary dilation with 1% tropicamide eye drops (Auromide eye drops, India) by a Retina specialist.

Anterior and posterior segment photographs were taken for records except in patients unable to cooperate for clinical pictures. Intraocular pressure (IOP) was recorded with Air-puff tonometer (Kowa KT-800) except in patients unable to rest chin on the slit lamp, in whom IOP was measured by tonopFen (I-Care).

The primary outcome was the presence of leukemia retinopathy, defined as any retinal abnormality attributable to leukemia (including intra-retinal hemorrhage, pre-retinal hemorrhage, Roth’s spots, cotton wool spot, or vascular changes). Because leukemia is a systemic disease, both eyes of each patient were examined independently. Analyses were conducted at the eye level, with each eye considered a separate unit of analysis. The presence of retinopathy in an eye was sufficient to classify that eye as positive for the outcome.

The current study demonstrates the factors associated with the development of retinopathy among acute and chronic Leukemia patients in Khmer-Soviet Friendship Hospital. Ocular involvement in leukemia is very frequent and 9% to 90% prevalence has been mentioned in various studies [8][9]. In this study, Leukemia retinopathy was detected in 28 eyes (60.86%) of 23 patients. The majority of retinal problems were intra-retinal hemorrhages 28 eyes (60.86%). Roth’s spots, the pathognomonic of Leukemia retinopathy represent 12 eyes (26.08%) of our study.

The prevalence of Leukemia retinopathy in present study is consistent with those reported by Koshy et al. [4], who documented leukemia retinopathy in 69% of patients and intra-retinal hemorrhage in 53.5% of cases, suggesting that retinal hemorrhages are a common sign in ocular involvement in leukemia (Table 3). Interestingly, our study demonstrated the highest prevalence of Roth’s spots (26.08%) among all reviewed studies. This may reflect differences in diagnostic sensitivity, patient demographics, or disease stage at presentation. Roth’s spots may occur in other systemic conditions such as bacterial endocarditis, anemia, or vascular diseases. However, in our study, patients with diabetes, hypertension, autoimmune disorders, vascular occlusions, or hematological disorders other than leukemia were excluded. Therefore, the high prevalence of Roth’s spots observed in our cross-sectional study is unlikely to be explained by other comorbid pathologies and more likely reflects the ocular vascular compromise associated with leukemia itself. Roth’s spots, although less frequently identified in other studies, ranging from 6.3% in Jacob Koshy’s study to 23% in A.M. Abu El-Asrar et al., remain a significant indicator of retinal vascular compromise in leukemia [4][10].

Cotton wool spots were mentioned in 10.86% of our patients, a rate that is intermediate compared to other studies. While A.M. Abu El-Asrar et al. [10] reported a higher prevalence (16%), other studies such as Dina N. Laimon et al. and Jihene Sayadi reported markedly lower rates (0.9% and 3.1%, respectively) [9][11]. The variability in cotton wool spot detection may be attributed to differences in imaging techniques or clinical criteria used for diagnosis (Table 3).

Overall, the prevalence of leukemia retinopathy in our study (60.86%) aligns closely with the findings of Jacob Koshy [4] and Jihene Sayadi [11], reinforcing the notion that retinal involvement is a common and clinically relevant manifestation of leukemia. The relatively high rates of intra-retinal hemorrhage and Roth’s spots observed in our study underscore the importance of routine ophthalmologic screening in leukemia patients, as early detection of retinal changes may provide insights into systemic disease activity and guide therapeutic interventions.

In our cross-sectional study of 23 leukemia patients, hematologic parameters revealed significant deviations from normal ranges, reflecting the systemic impact of the disease. Notably, leukocyte counts were elevated in the majority of patients, with 60.87% exhibiting leukocytosis. Hemoglobin levels were markedly reduced in nearly all patients, with 95.65% showing anemia. This high frequency highlights the common occurrence of marrow infiltration and impaired erythropoiesis in leukemia, leading to decreased oxygen-carrying capacity and associated clinical symptoms such as fatigue and pallor. Similarly, low platelet count (thrombocytopenia) was found in 91.30% of cases, which is indicative of a higher risk of bleeding problems and the suppression of megakaryocytic lineage. The low platelet count aligns with the high incidence of intra-retinal hemorrhages noted in our ophthalmologic findings (Table 3).

Previous studies [5][10][12] reported associations between anemia, thrombocytopenia, and retinal hemorrhages in patients with leukemia. In our study, we found similar patterns; however, these associations did not reach statistical significance due to the small sample size. Therefore, these findings should be considered exploratory.

These hematologic abnormalities including leukocytosis, anemia, and thrombocytopenia are characteristic of leukemia and contribute to both systemic and ocular manifestations. Although lower platelet counts were observed frequently in patients with retinal hemorrhages, this association was not statistically significant. As a result, our report should be interpreted as exploratory and do not provide sufficient evidence to confirm a causal role of thrombocytopenia in leukemia retinopathy.

Table 3. Retinal manifestation in Leukemia retinopathy.

The correlation of hematologic parameters to retinopathy revealed high level of leukocyte counts showed higher odds of intra retinal hemorrhage, indicating a possible relationship between leukocytosis and retinal vascular changes, though this was not statistically significant. In addition, low rate of hemoglobin and platelet count were observed more frequent in intra retinal hemorrhage; however, the results are not statistically significant. Because the limited of sample size, these observations should be interpreted as exploratory rather than confirmatory. Correlation of Roth’s spot in Leukemia retinopathy elevated leukocyte counts were associated with higher odds which suggest a potential link to retinopathy but the results are also not statistically significant.

In term of vision, 6 patients (26.09%) have severe visual impairment; however, only 3 patients (13.04%) reported subjective complaints of vision deterioration. Some patients with severe visual impairment may not recognize or report visual difficulties due to gradual adaptation with the fellow eye, systemic illness overshadowing ocular symptoms, or limited awareness of visual standards. Therefore, while objective testing revealed a higher prevalence of severe impairment, most patients were clinically asymptomatic from their own perspective. Dina N. Laimon et al. [9] also reported 8.6% complained of deterioration of vision, while 6.8% had impaired vision and 1.8% had complete visual loss. Reza Mirshahi et al. [5] also mentioned 10.6% patients complained of vision loss. These indicated, the visual disturbance.

Our study has several limitations. First of all, the sample size is too small (23 cases) cannot cover all possibilities of all Leukemia patients in the study and it severely restricts the statistical power of the study. This is also a reason that we replace Fisher’s Exact Test to calculate odd ratio for crude analysis to a standard logistic regression as planned in the research proposal. This minuscule sample size resulted from the hospital’s shifting, which left not much time for internal policies and documentations. Secondly, there is no bone marrow aspiration results to confirm a gold standard diagnosis in some patients which is not a strongly evidence of the diagnosis. However, the first-line diagnostic approach in developing countries, clinical signs, basic hematology test and peripheral blood smear also be the practical realities of Leukemia diagnosis. Thirdly, some patients are already received the treatment such as chemotherapy, blood transfusion or platelet administration before the eye’s examination. These treatments may have improved hematological parameters and alleviated ocular signs and symptoms. This phenomenon may be described as temporal ambiguity, a limitation inherent in cross-sectional studies as it prevents clear determination of whether retinopathy was directly attributable to leukemia or influenced by treatment effects. Consequently, the observed correlations between hematological parameters and retinopathy should be interpreted with caution. However, this issue can be addressed through the future cohort studies.

Based on the result of our finding, we believe that retinal manifestations of leukemia could be detected at initial presentations or relapse. While some manifestations are asymptomatic, others could affect a patient’s visual acuity and change their quality of life, or even the disease course. Intra-retinal hemorrhages were the most common reported in Leukemia patients. High level of leukocyte count (Leukocytosis), low level of hemoglobin (anemia), and low level of platelet (thrombocytopenia) are characteristic of leukemia and contribute to both systemic and ocular manifestations.

As study reported high leukocytes, low platelet and hemoglobin are likely risk to have intra-retinal hemorrhage, so cooperation between ophthalmologists and haemato-oncologists by conducting a blood test at each patient’s appointment could facilitate the early detection or recognition of retinal manifestations resulting from abnormal hematological parameters. This approach can significantly enhance patient outcomes by allowing for timely interventions. Moreover, we need further evaluation of a larger cohort of acute and chronic leukemia patients especially for survival analysis to set the record for the prognostic value of Leukemia retinopathy in such neoplasms.

This study highlights several important factors that influence the prognosis factor of Leukemia patients. To medical professionals, this research identifies the benefit of eyes examination in Leukemia patients by early detection of relapse condition of disease and early worsening. In addition, it is an input contribute to assists Oncologist and Hematologist in prompting interventions, which benefits in better prognosis for patients. To patients, this present study demonstrates that a complete eyes examination with pupil dilation is a standard procedure in Leukemia patients which is a non-invasive technique. Moreover, other eyes diseases could be diagnosed during this examination.

All authors read and approved the final version of the manuscript and agree to be accountable for all aspects of the work.

Krytha Tor: Conceptualization, study design, data collection, ophthalmic examination, data interpretation, manuscript drafting, and manuscript revision; Dyna Khuon: Study co-supervision, statistical analysis, data interpretation, and manuscript revision; Remy Tor: Data collection, literature review, and manuscript revision; Leakhena Or: Data collection and manuscript revision; Keanghak Taing: Data analysis and manuscript revision; Long Chhuor: Patient recruitment, facilitation of clinical examination, and manuscript revision; Chanvatanak Ly: Study supervision, methodology, data interpretation, and critical revision of the manuscript.

The authors used ChatGPT to assist in purpose, improving the grammar, language, and readability of the manuscript after reviewers’ comments. The AI tool was not used for data collection, data analysis, interpretation of results, or generation of scientific conclusions. All content was reviewed, verified, and approved by the authors, who take full responsibility for the accuracy and integrity of the manuscript.

Table S1. Association between variable, intra-retinal hemorrhage, Roth’s spot and Cotton wool spot in Leukemia retinopathy.

Note: OR = Odds Ratio, CI = Confidence Interval, statistically significant (p < 0.05), Analyses were conducted at the eye level; percentages represent the proportion of eyes with lesions.

Table S2. Associate variable with treatment status.

Note: n = number of participants, % = percentage, Chi-square test used; Only variables with p < 0.05 in bivariate analysis proceed to multivariate analysis.