Summary
Sickle cell disease (SCD), encompassing genotypes such as HbSS and HbSC, causes chronic haemolysis and microvascular occlusion, leading to organ damage. The retina is particularly vulnerable, often resulting in sickle cell retinopathy (SCR) or sickle cell maculopathy (SCM). The precise underlying mechanisms are unclear, though various factors are suggested to contribute. This study explored the role of whole blood viscosity and red blood cell (RBC) deformability in SCR and SCM. Adult HbSS (n = 34) and HbSC patients (n = 34) were offered an ophthalmic examination to determine SCR stage. A venous ethylenediaminetetraacetic acid (EDTA) sample was collected from each participant. Whole blood viscosity was measured using a Brookfield viscometer and RBC deformability was assessed using the Oxygenscan feature of the Laser Optical Rotational Red Cell Analyser as a function of the (varying) partial oxygen pressure. HbSC patients with proliferative sickle cell retinopathy (PSCR) had a lower delta elongation index (p = 0.012) and point of sickling (p = 0.002) than those without PSCR, suggesting that RBC sickling might not play a central role in the pathogenesis of PSCR in HbSC patients. Despite hyperviscosity being a commonly proposed mechanism, no associations were found between blood viscosity, SCR and SCM. These results point to alternative mechanisms contributing to SCR and SCM, highlighting the complexity and need for further research to fully understand the underlying factors.
Keywords: erythrocyte deformability, sickle cell disease, sickle cell maculopathy, sickle cell retinopathy, viscosity
Sickle cell retinopathy and maculopathy are common ocular complications of sickle cell disease. This study assessed the contribution of red blood cell characteristics and blood viscosity to their development. HbSC patients with proliferative retinopathy showed a lower delta elongation index and point of sickling, while no link was found between whole blood viscosity and retinal changes. These findings suggest that factors beyond viscosity and sickling may drive retinal damage in sickle cell disease.
INTRODUCTION
Sickle cell disease (SCD), encompassing genotypes such as HbSS and HbSC, is characterized by chronic haemolysis and microvascular occlusion, resulting in chronic inflammation, ischaemia and organ damage. 1 The retina is particularly sensitive to ischaemic damage, leading to sickle cell retinopathy (SCR) or sickle cell maculopathy (SCM).
SCR is classified into non‐proliferative sickle cell retinopathy (NPSCR) and proliferative sickle cell retinopathy (PSCR). PSCR involves neovascularisation, which can cause vitreous haemorrhage or retinal detachment, drastically impairing visual acuity. Previous studies demonstrated that compound heterozygous sickle‐haemoglobin (Hb) C disease (HbSC genotype) is associated with SCR (particularly PSCR), while this genotype is linked to less severe systemic disease than homozygous SCD (HbSS genotype). 2 , 3 The mechanism underlying the association between HbSC genotype and PSCR remains unclear, but many pathophysiological aspects have been suggested to contribute, including increased blood viscosity related to higher Hb levels and haematocrit. This can lead to stagnant blood flow, causing polymerization, sickling of red blood cells (RBCs), vascular obstruction, ischaemia and angiogenesis. 4 , 5 , 6
SCM affects the central retina (the macula) and includes abnormalities such as macular thinning, enlargement of the foveal avascular zone (FAZ) and lower vessel densities. These abnormalities differ from those in the peripheral retina and are detectable through spectral‐domain coherence tomography (SD‐OCT) and optical coherence tomography angiography (OCTA). 7 , 8 Unlike SCR, SCM has been associated with the HbSS genotype. 9 , 10 While the functional consequences of SCM remain unclear, scotomas, reduced contrast sensitivity and impaired colour vision have been reported in SCD patients with severely diminished vessel density on OCTA, even with normal visual acuity. 11
SCR results from ischaemic vascular disease and secondary angiogenesis, but the key factors in its pathogenesis remain unclear. 5 , 12 One area of interest is RBC deformability, as impaired deformability could affect their ability to navigate through the microvasculature, exacerbating ischaemic conditions. The Oxygenscan is a new method to determine RBC deformability ex vivo as a function of the partial oxygen pressure. 13 This approach is particularly relevant to SCR and SCM, as RBC deformability might impact retinal microcirculation. Understanding the pathophysiology may open new paths of intervention regarding retinopathy and may help identify risk factors for PSCR. This study aimed to explore the role of whole blood viscosity, RBC deformability, and hypoxia‐induced sickling in SCR and SCM.
METHODS
This single‐centre observational study was conducted in collaboration between the departments of Ophthalmology and Haematology of the Amsterdam University Medical Centres (UMC), Amsterdam, the Netherlands. The study was approved by the institutional review board and ethics committee of the Amsterdam UMC (ClinicalTrials.gov ID: NCT06396403) and followed the principles of the Declaration of Helsinki. All participants gave written informed consent.
Patients with HbSS or HbSC genotype from the adult outpatient SCD clinic were approached between June 2023 and June 2024. Patients using crizanlizumab, voxelotor or mitapivat were excluded.
Ophthalmic data
Ophthalmic examination included best‐corrected visual acuity (VA), slit lamp examination and dilated fundoscopy. Visual impairment was defined as VA < 20/25. Fluorescein angiography was only used in case of tentative diagnosis of PSCR or for patients with stage III disease or worse. Therefore, SCR stage was determined based on fundoscopic examination (instead of the Goldberg classification) as (1) either no signs of SCR or signs of NPSCR (sunburst lesions, salmon patches, arteriolar occlusions, peripheral anastomose without neovascularization), or (2) PSCR (with neovascularization). Eyes with previously resolved vitreous haemorrhage and/or treated retinal detachment were also classified as PSCR.
Macular SD‐OCT/OCTA scans were obtained with the Optovue RTVue XR Avanti instrument (Optovue Inc., Fremont, California, USA). For each eye, a 3 × 3 mm scan and a 6 × 6 mm scan were acquired. AngioAnalytics software (beta‐version 2016.200.037) was used for projection artefact removal and automated segmentation of the capillary plexuses. Images were excluded if the quality index was ≤4 or motion artefacts limited grading. To assess the presence of SCM, two separate parameters were used: the presence of FAZ irregularities on OCTA (e.g. enlargement or pathological areas of vascular loss) and the presence of macular thinning on SD‐OCT (defined as the presence of blue areas on the retinal thickness colour‐coded map in at least one of the standard Early Treatment Diabetic Retinopathy Study [ETDRS] macular subfields). Figure 1 illustrates a normal SD‐OCT (1A/1B) and OCTA scan (1C) on the left and an abnormal SD‐OCT (1D/1E) and OCTA scan (1F) on the right. The assessment of the scans was conducted by a junior researcher (RPB) and a senior retinal specialist (RMHD).
FIGURE 1.
Illustration of normal and abnormal spectral‐domain coherence tomography (SD‐OCT) and optical coherence tomography angiography (OCTA) scans. (A, B) A normal SD‐OCT scan (A) with a completely green Early Treatment Diabetic Retinopathy Study (ETDRS) circle (B), indicating normal macular thickness. An example of a normal OCTA scan is shown in (C), displaying the typical macular vascular structure with clearly defined white vessels and a round, well‐formed foveal avascular zone (FAZ) at the centre of the macula, indicative of normal retinal vasculature. (D, E) An SD‐OCT scan with macular thinning. The yellow arrows in (D) show the focal areas of macular thinning. (E) The corresponding ETDRS circle, with several blue segments indicating macular thinning. The OCTA scan shown in (F) demonstrates an irregularly enlarged border of the foveal avascular zone, indicating local vascular loss (yellow arrow).
Haematological data
Haematological characteristics were retrieved from electronic patient records, including Hb genotype, hydroxycarbamide (hydroxyurea) use, presence of microalbuminuria, avascular necrosis (AVN), acute chest syndrome (ACS), stroke, chronic ulcers, cholelithiasis and laboratory characteristics (Hb, creatinine, lactate dehydrogenase, bilirubin, reticulocytes). Laboratory values were no more than 6 months old and were, in most cases, within a few weeks to months of the ophthalmic examination and Oxygenscan/blood viscosity measurements. Microalbuminuria was defined as a ratio between urinary microalbumin (mg/L) and creatinine (mmol/L) of more than 3.5. AVN was confirmed by X‐ray or magnetic resonance imaging (MRI). ACS was defined as a new pulmonary infiltrate on X‐ray in combination with respiratory symptoms/chest pain for which medical treatment was necessary. 14 Stroke was defined as a diagnosis of clinical overt stroke confirmed by MRI/computed tomography. Chronic ulcers were defined as chronic ulceration of the lower leg, without another explanation, not healed within 14 days. 15 Cholelithiasis was defined as the presence of gallstones (on ultrasound) or a previous cholecystectomy.
A 6 mL ethylenediaminetetraacetic acid (EDTA) tube of venous blood was collected from each participant. Whole blood viscosity was measured at native haematocrit at 37°C with a shear rate increasing from 0 to 150 s−1 (with the first valid measuring point at 16.6 s−1) using the Brookfield LV DVII+CPE40 spindle viscometer (Brookfield Engineering, MA, USA). Haematocrit was measured using the ADVIA 2120i haematology analyser (Siemens Healthineers, Erlangen, Germany). The haematocrit‐to‐viscosity ratio (HVR), reflecting oxygen transport effectiveness of blood, was calculated as haematocrit (%)/whole blood viscosity (mPa.s). 16 RBC characteristics were evaluated within 24 h of sample collection using the Oxygenscan feature of the Laser Optical Rotational Red Cell Analyser (Lorrca, RR Mechatronics, Zwaag, the Netherlands) to determine RBC deformability as a function of the (varying) partial oxygen pressure (pO2). A detailed description of the Oxygenscan methodology has been published previously. 13 In short, 50 μL of the EDTA sample was mixed with 5 mL Iso‐Oxy (a polyvinylpyrrolidone solution mimicking the pH and osmolarity of whole blood) after standardizing the sample to a RBC count of 200 × 106. Two millilitres of the mixture was inserted via a Luer‐lock adapter into the device's cylinder gap, creating a shear stress of 30 Pa on the RBCs. Deoxygenation was induced by introducing nitrogen gas, followed by passive reoxygenation with ambient air. During this process, the elongation index (EI) of RBCs was calculated multiple times, representing their deformability upon deoxygenation and reoxygenation. The following output of the Oxygenscan was used for analysis: the point of sickling (PoS), the maximum EI (EImax), the minimum EI defined at a pO2 of 10 mmHg (EImin) and the delta EI. The PoS is the pO2 at which more than a 5% decrease in EI occurs, representing the point where RBC sickling starts. A lower PoS represents a more favourable RBC deformability. EImax is the maximum (baseline) deformability at normal oxygen pressure. EImin is the minimum deformability at 10 mmHg upon deoxygenation. Delta EI is the difference between EImax and EImin, representing the degree of RBC sickling upon deoxygenation.
Statistical analysis
Statistical analysis was performed using IBM SPSS Statistics (v28.0; Armonk, New York). Differences in counts or means between the groups were examined with the Chi‐squared test (or Fisher's exact test when suitable) for binary variables and unpaired Student's t‐test for continuous variables. In the case of non‐normal distribution, Mann–Whitney U test was used. p‐values <0.05 were considered significant. Reported p‐values are uncorrected due to the exploratory character of this study.
RESULTS
A total of 68 patients were included: 34 patients with HbSS genotype and 34 patients with HbSC genotype. Baseline characteristics (including stratification by genotype) are outlined in Table 1. The cohort comprised 31 male patients (45.6%) and 37 female patients (54.4%). The mean age was 33.8 ± 11.8 years and did not significantly differ between genotypes, but patients with PSCR were older than patients without PSCR (37.4 ± 12.0 vs. 31.5 ± 11.2, p = 0.044). PSCR was present in 26 patients (38.2%), who predominantly had HbSC genotype (p < 0.001). OCTA scans of sufficient quality were available for 63 patients (92.6%) and SD‐OCT scans were available for 64 patients (94.1%). FAZ abnormalities were present in 31/63 patients (49.2%) and macular thinning in 31/64 patients (48.4%). Hydroxyurea was used by 27 patients (39.7%), predominantly with HbSS genotype (p < 0.001). To facilitate easier interpretation of the results, further details on the Oxygenscan parameters, their significance and all relevant abbreviations are provided in Table S1.
TABLE 1.
Baseline characteristics and stratification by genotype.
| Total (n = 68) | HbSS (n = 34) | HbSC (n = 34) | p‐value | |
|---|---|---|---|---|
| Sex | ||||
| Male (n, %) | 31 (45.6%) | 16 (47.1%) | 15 (44.1%) | 0.808 a |
| Female (n, %) | 37 (54.4%) | 18 (52.9%) | 19 (55.9%) | |
| Age (mean ± SD) | 33.8 ± 11.8 | 33.6 ± 11.7 | 34.0 ± 12.0 | 0.879 b |
| Retinopathy status | ||||
| No PSCR (n, %) | 42 (61.8%) | 29 (85.3%) | 13 (38.2%) | <0.001 a |
| PSCR (n, %) | 26 (38.2%) | 5 (14.7%) | 21 (61.8%) | |
| Maculopathy | ||||
| FAZ abnormalities | 31/63 (49.2%) c | 15/31 (48.4%) | 16/32 (50%) | 0.898 a |
| Macular thinning | 31/64 (48.4%) d | 20/33 (60.6%) | 11/31 (35.5%) | 0.044 a |
| Ocular complications | ||||
| Vitreous haemorrhage (n, %) | 10 (14.7%) | 1 (2.9%) | 9 (26.5%) | 0.006 a |
| Retinal detachment (n, %) | 2 (2.9%) | 0 (0%) | 2 (5.9%) | NA |
| Retinal laser treatment (n, %) | 15 (22.1%) | 3 (8.8%) | 12 (35.3%) | 0.008 a |
| Hydroxyurea use (n, %) | 27 (39.7%) | 22 (64.7%) | 5 (14.7%) | <0.001 a |
| Chronic transfusion use (n, %) | 2 (2.9%) | 2 (5.9%) | 0 (0%) | NA |
| Alpha‐thalassaemia (n, %) | 2 (2.9%) | 2 (5.9%) | 0 (0%) | NA |
| Laboratory characteristics (mean ± SD) | ||||
| Hb (g/dL) | 10.3 (±2.0) | 9.0 (±1.3) | 11.6 (±1.8) | <0.001 b |
| HbF (%) | 7.1 (±7.8) | 11.6 (±8.7) | 2.6 (±2.7) | <0.001 f |
| Creatinine (μmol/L) | 67.6 (±19.0) | 61.1 (±19.3) | 74.2 (±16.5) | <0.001 f |
| LDH (U/L) | 340.7 (±141.1) | 401.8 (±164.5) | 279.6 (±75.4) | <0.001 f |
| Bilirubin (μmol/L) | 33.9 (±24.0) | 45.3 (±27.1) | 22.6 (±13.0) | <0.001 f |
| Reticulocytes (%) | 5.7 (±3.9) | 8.4 (±3.8) | 3.1 (±1.0) | <0.001 f |
| Other organ damage | ||||
| Microalbuminuria, n (%) | 15 (22.1%) | 11 (32.4%) | 4 (11.8%) | 0.041 a |
| AVN, n (%) | 7 (10.3%) | 3 (8.8%) | 4 (11.8%) | 1.000 e |
| ACS, n (%) | 13 (19.1%) | 9 (26.5%) | 4 (11.8%) | 0.123 a |
| Stroke, n (%) | 1 (1.5%) | 1 (2.9%) | 0 (0%) | NA |
| Chronic ulcers, n (%) | 1 (1.5%) | 1 (2.9%) | 0 (0%) | NA |
| Cholelithiasis, n (%) | 24 (35.3%) | 17 (50%) | 7 (20.6%) | 0.011 a |
Abbreviations: ACS, acute chest syndrome; AVN, avascular necrosis; FAZ, foveal avascular zone; Hb, haemoglobin; HbF, foetal haemoglobin; LDH, lactate dehydrogenase; NA, not applicable; NPSCR, non‐proliferative sickle cell retinopathy; PSCR, proliferative sickle cell retinopathy; SCR, sickle cell retinopathy; SD, standard deviation.
Pearson Chi‐Square.
Unpaired Student's t‐test.
Scans only available for 63 of 68 patients.
Scans only available for 64 of 68 patients.
Fisher's exact test.
Mann–Whitney U test.
Differences in Oxygenscan and blood viscosity parameters between HbSS and HbSC genotype are outlined in Table 2. The EImin (p = 0.014) was higher in HbSC patients compared to HbSS patients, while the delta EI and PoS were lower (p < 0.001). Whole blood viscosity (p < 0.001) and haematocrit (p < 0.001) were also higher in HbSC patients compared to those in HbSS patients. Differences in EImax and HVR were not statistically significant.
TABLE 2.
Comparison of Oxygenscan/viscosity parameters between HbSS and HbSC genotypes.
| HbSS (n = 34) | HbSC (n = 34) | p‐value a | |
|---|---|---|---|
| EImax (mean ± SD) | 0.448 ± 0.076 | 0.417 ± 0.086 | 0.121 |
| EImin (mean ± SD) | 0.216 ± 0.130 | 0.293 ± 0.122 | 0.014 |
| Delta EI (mean ± SD) | 0.232 ± 0.085 | 0.124 ± 0.074 | <0.001 |
| PoS5% (mmHg; mean ± SD) | 36.9 ± 12.6 | 21.2 ± 6.4 | <0.001 |
| Whole blood viscosity (mPa.s; mean ± SD) | 4.1 ± 0.6 b | 5.2 ± 1.0 | <0.001 |
| Haematocrit (L/L, mean ± SD) | 0.288 ± 0.043 b | 0.351 ± 0.045 | <0.001 |
| HVR (mean ± SD) | 7.2 ± 0.9 b | 6.9 ± 0.8 | 0.227 |
Abbreviations: Delta EI, the difference between EImax and EImin (representing the degree of RBC sickling upon deoxygenation); EImax, the maximum (baseline) deformability at normal oxygen pressure; EImin, the minimum deformability at 10 mmHg upon deoxygenation; HVR, haematocrit‐to‐viscosity ratio (reflects the oxygen transport effectiveness of blood); pO2, partial oxygen pressure; PoS5%, point of sickling at 5% decrease of the deformability (representing the pO2 where RBC sickling starts); RBC, red blood cell; SD, standard deviation.
Unpaired Student's t‐test.
Analysis of 32 patients since measurement failed in two patients (one due to clotting and one due to paucity of material).
HbSS genotype
Table 3 demonstrates the differences in Oxygenscan and blood viscosity parameters between patients with and without PSCR, displayed for both genotypes separately. Oxygenscan and RBC rheology parameters were not significantly different between HbSS patients with PSCR and without PSCR.
TABLE 3.
Comparison of Oxygenscan/viscosity parameters between retinopathy groups for HbSS and HbSC genotypes.
| HbSS genotype | HbSC genotype | |||||
|---|---|---|---|---|---|---|
| No PSCR (n = 29) | PSCR (n = 5) | p‐value a | No PSCR (n = 13) | PSCR (n = 21) | p‐value a | |
| EImax (mean ± SD) | 0.445 ± 0.077 | 0.465 ± 0.076 | 0.607 | 0.414 ± 0.058 | 0.419 ± 0.101 | 0.847 |
| EImin (mean ± SD) | 0.208 ± 0.128 | 0.263 ± 0.147 | 0.391 | 0.243 ± 0.101 | 0.325 ± 0.126 | 0.057 |
| Delta EI (mean ± SD) | 0.238 ± 0.085 | 0.202 ± 0.090 | 0.398 | 0.171 ± 0.091 | 0.094 ± 0.043 | 0.012 |
| PoS5% (mmHg; mean ± SD) | 37.4 ± 12.9 | 33.9 ± 11.6 | 0.576 | 26.0 ± 6.9 | 18.2 ± 3.9 | 0.002 |
| Whole blood viscosity (mPa.s; mean ± SD) | 4.1 ± 0.7 b | 3.8 ± 0.3 | 0.336 | 5.3 ± 1.0 | 5.1 ± 1.1 | 0.531 |
| Haematocrit (L/L, mean ± SD) | 0.288 ± 0.043 b | 0.293 ± 0.050 | 0.800 | 0.351 ± 0.045 | 0.352 ± 0.046 | 0.956 |
| HVR (mean ± SD) | 7.1 ± 1.0 b | 7.7 ± 0.8 | 0.209 | 6.7 ± 0.6 | 7.1 ± 0.9 | 0.193 |
Abbreviations: Delta EI, the difference between EImax and EImin (representing the degree of RBC sickling upon deoxygenation); EImax, the maximum (baseline) deformability at normal oxygen pressure; EImin, the minimum deformability at 10 mmHg upon deoxygenation; HVR, haematocrit‐to‐viscosity ratio (reflects the oxygen transport effectiveness of blood); pO2, partial oxygen pressure; PoS5%, point of sickling at 5% decrease of the deformability (representing the pO2 where RBC sickling starts); PSCR, proliferative sickle cell retinopathy; RBC, red blood cell; SD, standard deviation.
Unpaired Student's t‐test.
Analysis of 27 patients since measurement failed in two patients (one due to clotting and one due to paucity of material).
For patients with HbSS, SD‐OCT scans were available for 33/34 patients and OCTA scans for 31/34 patients. The findings in Oxygenscan and blood viscosity parameters between patients with and without FAZ abnormalities and macular thinning are displayed in Table 4. No significant correlations were found between the parameters and FAZ abnormalities and macular thinning. However, after stratification by sex, the PoS was lower in male HbSS patients with FAZ abnormalities compared to those without FAZ abnormalities (36.7 ± 11.8 vs. 50.7 ± 11.9, p = 0.043).
TABLE 4.
Associations between Oxygenscan/viscosity parameters and SCM for HbSS genotype.
| FAZ abnormalities | Macular thinning | |||||
|---|---|---|---|---|---|---|
| No (n = 16) | Yes (n = 15) | p‐value a | No (n = 13) | Yes (n = 20) | p‐value a | |
| EImax (mean ± SD) | 0.452 ± 0.075 | 0.427 ± 0.071 | 0.339 | 0.448 ± 0.086 | 0.447 ± 0.073 | 0.977 |
| EImin (mean ± SD) | 0.206 ± 0.121 | 0.176 ± 0.089 | 0.448 | 0.235 ± 0.154 | 0.196 ± 0.112 | 0.400 |
| Delta EI (mean ± SD) | 0.246 ± 0.074 | 0.250 ± 0.070 | 0.878 | 0.213 ± 0.099 | 0.251 ± 0.071 | 0.201 |
| PoS5% (mmHg; mean ± SD) | 39.2 ± 13.1 | 37.6 ± 11.3 | 0.721 | 34.9 ± 13.4 | 39.0 ± 12.0 | 0.365 |
| Whole blood viscosity (mPa.s; mean ± SD) | 4.0 ± 0.7 | 4.2 ± 0.6 | 0.464 | 4.0 ± 0.6 | 4.1 ± 0.7 | 0.761 |
| Haematocrit (L/L, mean ± SD) | 0.281 ± 0.046 | 0.297 ± 0.042 | 0.324 | 0.293 ± 0.046 | 0.287 ± 0.044 | 0.679 |
| HVR (mean ± SD) | 7.1 ± 1.11 | 7.2 ± 0.8 | 0.908 | 7.3 ± 0.9 | 7.1 ± 1.0 | 0.508 |
Abbreviations: Delta EI, the difference between EImax and EImin (representing the degree of RBC sickling upon deoxygenation); EImax, the maximum (baseline) deformability at normal oxygen pressure; EImin, the minimum deformability at 10 mmHg upon deoxygenation; FAZ abnormalities, abnormalities of the foveal avascular zone; HVR, haematocrit‐to‐viscosity ratio (reflects the oxygen transport effectiveness of blood); pO2, partial oxygen pressure; PoS5%, point of sickling at 5% decrease of the deformability (representing the pO2 where RBC sickling starts); RBC, red blood cell; SCM, sickle cell maculopathy; SD, standard deviation.
Unpaired Student's t‐test.
HbSC genotype
In HbSC patients, the mean PoS was lower in patients with PSCR compared to patients without PSCR (18.2 ± 3.9 vs. 26.0 ± 6.9, p = 0.002), indicating that their RBCs are less prone to sickling under hypoxic conditions. Delta EI was also lower in patients with PSCR (0.094 ± 0.043 vs. 0.171 ± 0.091, p = 0.012), as shown in Table 3. Stratified analysis by sex revealed that the difference in PoS was significant for both male and female patients (p = 0.004 and p = 0.039 respectively), while the difference in Delta EI was significant only in male patients (p < 0.001). No differences were found for EImax and EImin between HbSC patients with and without PSCR. However, after stratifying by sex, a significant difference in EImin was observed, with male HbSC patients with PSCR having a higher mean EImin compared to those without PSCR (0.331 ± 0.128 vs. 0.183 ± 0.055, p = 0.014). Whole blood viscosity, haematocrit and HVR were not different between patients with and without PSCR.
SD‐OCT scans were available for 31/34 HbSC patients and OCTA scans were available for 32/34 patients. Table 5 demonstrates the differences in Oxygenscan and blood viscosity parameters between patients with and without FAZ abnormalities and macular thinning. Patients with FAZ abnormalities showed a lower EImax compared to patients without FAZ abnormalities (p = 0.017). After stratification by sex, this difference was found to be significant only in male patients (p = 0.026). Other Oxygenscan parameters, whole blood viscosity, haematocrit and HVR, were not different between patients with and without FAZ abnormalities and macular thinning.
TABLE 5.
Associations between Oxygenscan/viscosity parameters and SCM for HbSC genotype.
| FAZ abnormalities | Macular thinning | |||||
|---|---|---|---|---|---|---|
| No (n = 16) | Yes (n = 16) | p‐value a | No (n = 20) | Yes (n = 11) | p‐value a | |
| EImax (mean ± SD) | 0.447 ± 0.074 | 0.376 ± 0.083 | 0.017 | 0.421 ± 0.082 | 0.389 ± 0.094 | 0.331 |
| EImin (mean ± SD) | 0.317 ± 0.124 | 0.246 ± 0.097 | 0.080 | 0.302 ± 0.123 | 0.238 ± 0.094 | 0.142 |
| Delta EI (mean ± SD) | 0.129 ± 0.088 | 0.130 ± 0.056 | 0.971 | 0.119 ± 0.079 | 0.152 ± 0.058 | 0.244 |
| PoS5% (mmHg; mean ± SD) | 21.5 ± 6.4 | 22.0 ± 6.2 | 0.822 | 21.3 ± 6.3 | 23.0 ± 6.1 | 0.471 |
| Whole blood viscosity (mPa.s; mean ± SD) | 4.9 ± 0.8 | 5.5 ± 1.1 | 0.134 | 5.2 ± 1.1 | 5.3 ± 0.9 | 0.627 |
| Haematocrit (L/L, mean ± SD) | 0.346 ± 0.047 | 0.359 ± 0.044 | 0.424 | 0.347 ± 0.047 | 0.364 ± 0.043 | 0.345 |
| HVR (mean ± SD) | 7.1 ± 0.7 | 6.7 ± 0.9 | 0.187 | 6.9 ± 0.9 | 6.9 ± 0.7 | 0.950 |
Abbreviations: Delta EI, the difference between EImax and EImin (representing the degree of RBC sickling upon deoxygenation); EImax, the maximum (baseline) deformability at normal oxygen pressure; EImin, the minimum deformability at 10 mmHg upon deoxygenation; FAZ abnormalities, abnormalities of the foveal avascular zone; HVR, haematocrit‐to‐viscosity ratio (reflects the oxygen transport effectiveness of blood); pO2, partial oxygen pressure; PoS5%, point of sickling at 5% decrease of the deformability (representing the pO2 where RBC sickling starts); RBC, red blood cell; SCM, sickle cell maculopathy; SD, standard deviation.
Unpaired Student's t‐test.
DISCUSSION
SCR and SCM can impair visual function, but their underlying pathogenesis remains unclear. Suggested mechanisms include higher haematocrit, increased blood viscosity and vaso‐occlusion caused by RBC sickling (see Figure 2). This study evaluated the association between whole blood viscosity, haematocrit, RBC deformability, hypoxia‐induced sickling and SCR/SCM. No association was found between haematocrit or blood viscosity and PSCR, contradicting the hypothesis that these factors contribute to its higher incidence in HbSC patients. With respect to rheological characteristics, HbSC patients with PSCR had a lower PoS and delta EI compared to HbSC patients without PSCR, with the difference in delta EI being significant only in male patients. This indicates that other factors than early sickling at higher oxygen tension may play a key role in PSCR pathogenesis.
FIGURE 2.
Hypothesised pathophysiology of sickle cell retinopathy and maculopathy. This diagram illustrates the sequence of events leading to sickle cell retinopathy and maculopathy, highlighting the complex interactions between sickling, blood viscosity and retinal damage. It depicts how these factors are hypothesized to contribute to the development and progression of the conditions.
RBC deformability in SCR and SCM has been studied, but not using the Oxygenscan method. One study among paediatric HbSS patients evaluated RBC deformability as a function of shear stress, demonstrating that temporal macular thinning was associated with a decreased EI (representing the mean RBC deformability) at a shear stress of ≥3 Pa. 17 In adults with SCD, no associations between EI (RBC deformability under normoxic conditions) and SCR/SCM were found. 18 , 19 The Oxygenscan offers distinct advantages over earlier methods by measuring RBC deformability under varying oxygen conditions instead of varying shear stress. This is especially valuable in SCD as deoxygenation causes RBC sickling. Therefore, this method is expected to provide deeper insights into RBC deformability and sickling.
The PoS (the pO2 at which the EI decreases by 5%) marks the start of RBC sickling. A higher PoS reflects a greater sickling propensity. A previous study showed that the PoS is higher in HbSS patients than in those with other genotypes and lower in patients receiving treatment (e.g. voxelotor and transfusion therapy). 13 The delta EI indicates the proportion of RBCs that sickle upon deoxygenation and is, together with EImax, an indicator of anti‐sickling therapy effectiveness. 13 A lower delta EI indicates fewer sickling RBCs. A previous study demonstrated that the delta EI was lower in HbSC patients compared to HbSS patients and lower in HbSS patients who received transfusions compared to HbSS patients who did not. 20 Interestingly, our study showed that the PoS and delta EI were lower in HbSC patients with PSCR than in HbSC patients without PSCR, indicating that their RBCs are less prone to sickle under hypoxic conditions. This finding aligns with the observation that HbSC patients typically have less severe systemic disease, yet suggests that RBC sickling may not be a major factor in PSCR pathogenesis among HbSC patients. Regarding SCM, the delta EI presents a contrasting pattern, as patients with macular thinning showed a tendency for higher PoS and delta EI, indicating more pronounced RBC sickling upon deoxygenation. This aligns with previous studies showing SCM is more frequently associated with HbSS genotype characteristics (i.e. vaso‐occlusion). 21 , 22 , 23 However, the higher delta EI trend in patients with macular thinning in our study did not reach statistical significance. This lack of significance may suggest that while the trend aligns with the understanding of SCM pathogenesis, the effect of RBC sickling on SCM in our cohort may not be as pronounced as previously hypothesized. Larger studies are needed to validate these findings and clarify the relationship between RBC deformability and SCR/SCM.
The correlation between RBC deformability and SCR/SCM was predominantly observed in male patients. This could be due to the protective effects of oestrogen on endothelial function, which may mitigate sickling‐related vascular damage. 2 , 24 Male patients, lacking this protection, may be more susceptible to impaired RBC deformability. Higher Hb levels in males may also contribute, 25 though our study found no differences in Hb levels between males with and without retinal abnormalities, suggesting other factors are involved. Male HbSS patients with FAZ abnormalities had lower PoS (and therefore lower RBC sickling) compared to those without FAZ abnormalities. This contradicts the hypothesis that a higher degree of sickling causes ischaemic macular damage. Other factors (such as cell membrane damage from different causes) might also result in lower PoS and contribute to disease severity. However, given the small sample size and multiple variables, this finding could be coincidental.
The results suggest that mechanisms beyond RBC sickling contribute to PSCR pathogenesis. One hypothesis is blood hyperviscosity, though earlier studies show contradictory results: some found an association between higher blood viscosity and PSCR, 18 , 26 while others did not. 19 , 27 Our study found no difference in whole blood viscosity between patients with and without PSCR. Differences between studies may arise from factors like the dynamic nature of blood viscosity (especially in patients receiving blood transfusions). Unlike blood with normal Hb, blood viscosity in SCD can fluctuate depending on oxygen levels, complicating accurate measurements. Hypoxia influences blood rheology (including whole blood viscosity) in SCD, and a previous study suggests that the hypoxia‐dependent increase of whole blood viscosity might lead to lower flow velocities, increased RBC sickling, vaso‐occlusion, ischaemia and additional increase in blood viscosity. 28 Previous studies, including ours, measured blood viscosity under normoxia, which may not fully capture hypoxia‐related changes in viscosity, potentially contributing to contrasting findings between studies. Hydroxyurea use, which alters RBC rheological properties, might also contribute. However, an exploratory analysis (data not shown) excluding hydroxyurea‐treated patients showed similar results, with no significant differences in blood viscosity. Further research is needed to explore this aspect in more detail. Phlebotomy reduces blood viscosity and therefore hyperviscosity‐related organ complications in SCD, making it a potential preventive treatment for PSCR, given the hypothesized role of blood hyperviscosity in its pathophysiology. 29 , 30 However, the absence of an association between blood viscosity and PSCR could suggest that the benefit of phlebotomy might be limited in PSCR. Furthermore, the differences in HVR (reflecting oxygen transport effectiveness of blood) were not significant. Exploring other potential mechanisms, such as endothelial dysfunction and angiogenesis, could offer a deeper understanding of PSCR.
This study has several limitations. The low number of HbSS patients with PSCR may affect the robustness of our findings. Additionally, the prevalence of Goldberg stage I/II retinopathy might be underestimated, as fluorescein angiography was not routinely performed. 31 The Oxygenscan measurements were performed within 24 h of blood sampling, which is a potential limitation, though the method is validated for this time frame. 32 Another limitation is that whole blood viscosity can fluctuate with oxygen levels in SCD, complicating accurate measurements, since low oxygen levels influence blood rheology. 28 We also chose to keep p‐values uncorrected due to the exploratory character of this study, which might increase the error rate with multiple analyses.
In conclusion, this study suggests that RBC sickling does not play a central role in the pathogenesis of PSCR in HbSC patients. Although blood hyperviscosity is often proposed as an underlying mechanism, our study did not find any associations between blood viscosity, SCR and SCM. These results point to alternative mechanisms contributing to SCR and SCM development, highlighting the complexity of these conditions and the need for further research to fully understand the underlying factors.
AUTHOR CONTRIBUTIONS
This study was designed by BJB and RMHD. Sample and data collection was performed by RPB, MV and HK. Initial analysis was performed by RPB and MV and critically reviewed by RMHD, IK, ROS, EN and BJB. The manuscript was drafted by RPB and critically revised by RMHD, IK, MV, HK, RVZ, ROS, EN and BJB. All authors approved the final version.
FUNDING INFORMATION
This work was supported by Stichting Pupil (Pupil Foundation, Amsterdam, the Netherlands), Landelijke Stichting voor Blinden en Slechtzienden, the Oogfonds and Stichting Beheer het Schild through Stichting UitZicht (UitZicht Foundation, Ede, The Netherlands, UZ 2022‐24), Het Sikkelcelfonds (www.hetsikkelcelfonds.nl, the Netherlands), the Maarten Kappelle Stichting (the Netherlands) and the Rotterdamse Stichting Blindenbelangen (the Netherlands). The funding organizations had no role in the design or conduct of this research.
CONFLICT OF INTEREST STATEMENT
RPB has received research grants from Stichting Pupil (Pupil Foundation, the Netherlands), Stichting UitZicht (UitZicht Foundation, the Netherlands), Het Sikkelcelfonds (the Netherlands), the Maarten Kappelle Stichting (the Netherlands) and the Rotterdamse Stichting Blindenbelangen (the Netherlands). ROS has received research grants from Novartis and Boeringer‐Ingelheim and participated in advisory board meetings of Apellis, Boeringer‐Ingelheim and Ciana Therapeutics. EN has received a research grant from Novartis and participated in the advisory board and speakers' bureau of Novartis. BJB has received research grants from Sanquin, Novartis, GBT/Pfizer and BMS/Celgene; participated in advisory board meetings of BMS/Celgene, Novo Nordisk and GBT/Pfizer; and received honoraria from Sanofi and Novo Nordisk. RMHD and IK have no conflicts of interest to declare.
ETHICS APPROVAL STATEMENT
The study was approved by the institutional review board and ethics committee of the Amsterdam University Medical Centres and carried out in accordance with the principles of the Declaration of Helsinki (seventh revision, 2013).
PATIENT CONSENT STATEMENT
All participants gave written informed consent before participation.
CLINICAL TRIAL REGISTRATION
Supporting information
Table S1.
ACKNOWLEDGEMENTS
We thank Stichting Pupil (Pupil Foundation), Landelijke Stichting voor Blinden en Slechtzienden, the Oogfonds and Stichting Beheer het Schild through Stichting UitZicht (UitZicht Foundation), Het Sikkelcelfonds, theMaarten Kappelle Stichting and the Rotterdamse Stichting Blindenbelangen for supporting this work by their grants.
Brandsen RP, Diederen RMH, Klaassen I, Veldthuis M, Korsten H, van Zwieten R, et al. The role of red blood cell characteristics and viscosity in sickle cell retinopathy and maculopathy. Br J Haematol. 2025;206(6):1796–1805. 10.1111/bjh.20124
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
- 1. Conran N, Belcher JD. Inflammation in sickle cell disease. Clin Hemorheol Microcirc. 2018;68(2–3):263–299. 10.3233/CH-189012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Brandsen RP, Diederen RMH, Bakhlakh S, Nur E, Schlingemann RO, Biemond BJ. Natural history and rate of progression of retinopathy in adult patients with sickle cell disease: an 11‐year follow‐up study. Blood Adv. 2023;7(13):3080–3086. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Duan XJ, Lanzkron S, Linz MO, Ewing C, Wang J, Scott AW. Clinical and ophthalmic factors associated with the severity of sickle cell retinopathy. Am J Ophthalmol. 2019;197:105–113. [DOI] [PubMed] [Google Scholar]
- 4. Abdalla Elsayed MEA, Mura M, Al Dhibi H, Schellini S, Malik R, Kozak I, et al. Sickle cell retinopathy. A focused review. Graefes Arch Clin Exp Ophthalmol. 2019;257(7):1353–1364. 10.1007/s00417-019-04294-2 [DOI] [PubMed] [Google Scholar]
- 5. Andrawes NG, Ismail EA, Roshdy MM, Ebeid FSE, Eissa DS, Ibrahim AM. Angiopoietin‐2 as a marker of retinopathy in children and adolescents with sickle cell disease: relation to subclinical atherosclerosis. J Pediatr Hematol Oncol. 2019;41(5):361–370. [DOI] [PubMed] [Google Scholar]
- 6. Darbari DS, Onyekwere O, Nouraie M, Minniti CP, Luchtman‐Jones L, Rana S, et al. Markers of severe vaso‐occlusive painful episode frequency in children and adolescents with sickle cell anemia. J Pediatr. 2012;160(2):286–290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Minvielle W, Caillaux V, Cohen SY, Chasset F, Zambrowski O, Miere A, et al. Macular Microangiopathy in sickle cell disease using optical coherence tomography angiography. Am J Ophthalmol. 2016;164:137–144.e1. [DOI] [PubMed] [Google Scholar]
- 8. Hoang QV, Chau FY, Shahidi M, Lim JI. Central macular splaying and outer retinal thinning in asymptomatic sickle cell patients by spectral‐domain optical coherence tomography. Am J Ophthalmol. 2011;151(6):990–994.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Lim JI, Cao D. Analysis of retinal thinning using spectral‐domain optical coherence tomography Imaging of sickle cell retinopathy eyes compared to age‐ and race‐matched control eyes. Am J Ophthalmol. 2018;192:229–238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Hussnain SA, Coady PA, Slade MD, Carbonella J, Pashankar F, Adelman RA, et al. Hemoglobin level and macular thinning in sickle cell disease. Clin Ophthalmol. 2019;13:627–632. 10.2147/OPTH.S195168 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Martin GC, Dénier C, Zambrowski O, Grévent D, Bruère L, Brousse V, et al. Visual function in asymptomatic patients with homozygous sickle cell disease and temporal macular atrophy. JAMA Ophthalmol. 2017;135(10):1100–1105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Duits AJ, Rodriguez T, Schnog JJ. Serum levels of angiogenic factors indicate a pro‐angiogenic state in adults with sickle cell disease. Br J Haematol. 2006;134(1):116–119. 10.1111/j.1365-2141.2006.06103.x [DOI] [PubMed] [Google Scholar]
- 13. Rab MAE, van Oirschot BA, Bos J, Merkx TH, van Wesel ACW, Abdulmalik O, et al. Rapid and reproducible characterization of sickling during automated deoxygenation in sickle cell disease patients. Am J Hematol. 2019;94(5):575–584. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Stuart MJ, Setty BN. Sickle cell acute chest syndrome: pathogenesis and rationale for treatment. Blood. 1999;94(5):1555–1560. [PubMed] [Google Scholar]
- 15. Koshy M, Entsuah R, Koranda A, Kraus AP, Johnson R, Bellvue R, et al. Leg ulcers in patients with sickle cell disease. Blood. 1989;74(4):1403–1408. [PubMed] [Google Scholar]
- 16. Waltz X, Hardy‐Dessources M‐D, Lemonne N, Mougenel D, Lalanne‐Mistrih M‐L, Lamarre Y, et al. Is there a relationship between the hematocrit‐to‐viscosity ratio and microvascular oxygenation in brain and muscle? Clin Hemorheol Microcirc. 2015;59:37–43. [DOI] [PubMed] [Google Scholar]
- 17. Martin GC, Brousse V, Connes P, Grevent D, Kossorotoff M, Da Costa L, et al. Retinal atrophy and markers of systemic and cerebrovascular severity in homozygous sickle cell disease. Eur J Ophthalmol. 2022;32(6):3258–3266. [DOI] [PubMed] [Google Scholar]
- 18. Lemaire C, Lamarre Y, Lemonne N, Waltz X, Chahed S, Cabot F, et al. Severe proliferative retinopathy is associated with blood hyperviscosity in sickle cell hemoglobin‐C disease but not in sickle cell anemia. Clin Hemorheol Microcirc. 2013;55(2):205–212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Beral L, Lemonne N, Romana M, Charlot K, Billaud M, Acomat M, et al. Proliferative retinopathy and maculopathy are two independent conditions in sickle cell disease: is there a role of blood rheology? Clin Hemorheol Microcirc. 2019;71(3):337–345. 10.3233/CH-180412 [DOI] [PubMed] [Google Scholar]
- 20. Boisson C, Rab MAE, Nader E, Renoux C, Kanne C, Bos J, et al. Effects of genotypes and treatment on oxygenscan parameters in sickle cell disease. Cells. 2021;10(4):811. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Hoyek S, Lemire C, Halawa O, Altamirano‐Lamarque F, Gonzalez E, Patel NA. Longitudinal assessment of macular thickness and microvascular changes in children with sickle cell disease. Ophthalmol Retina. 2024;8(2):184–194. [DOI] [PubMed] [Google Scholar]
- 22. Fares S, Hajjar S, Romana M, Connes P, Acomat M, Zorobabel C, et al. Sickle cell maculopathy: microstructural analysis using OCTA and identification of genetic, systemic, and biological risk factors. Am J Ophthalmol. 2021;224:7–17. [DOI] [PubMed] [Google Scholar]
- 23. Ong SS, Linz MO, Li X, Liu TYA, Han IC, Scott AW. Retinal thickness and microvascular changes in children with sickle cell disease evaluated by optical coherence tomography (OCT) and OCT angiography. Am J Ophthalmol. 2020;209:88–98. [DOI] [PubMed] [Google Scholar]
- 24. Gladwin MT, Schechter AN, Ognibene FP, Coles WA, Reiter CD, Schenke WH, et al. Divergent nitric oxide bioavailability in men and women with sickle cell disease. Circulation. 2003;107(2):271–278. [DOI] [PubMed] [Google Scholar]
- 25. Serjeant BE, Mason KP, Acheson RW, Maude GH, Stuart J, Serjeant GR. Blood rheology and proliferative retinopathy in homozygous sickle cell disease. Br J Ophthalmol. 1986;70(7):522–525. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Ranque B, Diaw M, Dembele AK, Lapoumeroulie C, Offredo L, Tessougue O, et al. Association of haemolysis markers, blood viscosity and microcirculation function with organ damage in sickle cell disease in sub‐Saharan Africa (the BIOCADRE study). Br J Haematol. 2023;203(2):319–326. [DOI] [PubMed] [Google Scholar]
- 27. Serjeant BE, Mason KP, Condon PI, Hayes RJ, Kenny MW, Stuart J, et al. Blood rheology and proliferative retinopathy in sickle cell‐haemoglobin C disease. Br J Ophthalmol. 1984;68(5):325–328. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Kucukal E, Man Y, Hill A, Liu S, Bode A, An R, et al. Whole blood viscosity and red blood cell adhesion: potential biomarkers for targeted and curative therapies in sickle cell disease. Am J Hematol. 2020;95(11):1246–1256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Padaro E, Kueviakoe IMD, Agbétiafa K, Magnang H, Mawussi K, Layibo Y, et al. Therapeutic phlebotomy during major sickle cell disease in Togo. Med Sante Trop. 2019;29(1):106–107. [DOI] [PubMed] [Google Scholar]
- 30. Lionnet F, Hammoudi N, Stojanovic KS, Avellino V, Grateau G, Girot R, et al. Hemoglobin sickle cell disease complications: a clinical study of 179 cases. Haematologica. 2012;97(8):1136–1141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Han IC, Zhang AY, Liu TYA, Linz MO, Scott AW. Utility of ultra‐widefield retinal imaging for the staging and management of sickle cell retinopathy. Retina. 2019;39(5):836–843. [DOI] [PubMed] [Google Scholar]
- 32. Rab MAE, Kanne CK, Bos J, Boisson C, van Oirschot BA, Nader E, et al. Methodological aspects of the oxygenscan in sickle cell disease: a need for standardization. Am J Hematol. 2020;95(1):E5–e8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.