Patient Selection

All patients had electrophoretic confirmation of sickle cell disease (SS), sickle cell–hemoglobin C disease (SC), or sickle-cell thalassemia (S-thal). Medical histories and sickle hemoglobinopathy type were recorded from the medical record. Initial evaluation included best-corrected visual acuity (BCVA) measurement by using the Snellen chart, Goldmann applanation tonometry, slit-lamp biomicroscopy, and dilated fundus examination. Staging of sickle retinopathy was based on the classification developed by Goldberg (stage I: peripheral arteriolar occlusion; stage II: arteriovenous anastomoses; stage III: peripheral neovascularization; stage IV: presence of vitreous hemorrhage; stage V: tractional and/or rhegmatogenous retinal detachment). All patients were staged by the same observer (J.I.L.). Patients with lens or other ocular media opacities, extensive laser or cryotherapy treatment, diabetes mellitus, uncontrolled systemic hypertension, or clinical evidence of any other maculopathies were excluded from the current study.

SD-OCT Retinal Thickness Mapping

SD-OCT was performed using the Spectral OCT/SLO system (OPKO Instrumentations) to obtain both OCT and SLO images with an axial resolution of <10 μm and a transverse resolution of 20 μm (in tissue). The system uses an SLO fundus image for alignment, orientation, and registration of the OCT image topographic maps. Both the Line scan (B-scan) and the 3-Dimension Retinal Topography scan protocols were used for image acquisition. The Line scan mode allows the capture of cross-sectional B-scan OCT images of the vitreoretinal, retinal, and chorioretinal structures. A red scanning line on the SLO image represents the exact location of the cross-sectional OCT image. We used the “Max Frame Count” of 32 frames. The “Max Frame Count” is the maximum sequentially captured frames of OCT and SLO images, which are captured and displayed as individual frames. The 3D Retinal Topography mode covers an area of 9.0 × 9.0 mm with a 2.0-mm depth. The retinal thickness map is displayed as 9 Early Treatment Diabetic Retinopathy Study (ETDRS)-like subfields that correspond to the microperimetry testing area, including central, and parafoveal and perifoveal superior, temporal, inferior, and nasal, subfields ( Figure 1 ). The central subfield included the circle centered on the fovea with a radius of 0.5 mm. The parafoveal subfields included the concentric ring of retina around the central subfield with an inner radius of 0.5 mm and an outer radius of 1.5 mm from the fovea. The perifoveal subfields included the outer ring of retina beyond the parafoveal subfields concentric with the fovea and with an inner radius of 1.5 mm and an outer radius of 3 mm. Focal macular thinning was defined as abrupt asymmetric decrease in retinal thickness confirmed on thickness map and B-scan.

Representative macular spectral-domain optical coherence tomography (SD-OCT) maps showing mean thicknesses of 9 ETDRS-like subfields within the Polar 3 microperimetry testing grid. (Left) SD-OCT map of right eye of Patient 3 (SS sickle cell patient, best-corrected visual acuity 20/25), demonstrating significant perifoveal and parafoveal temporal thinning and therefore grouped to Group A (“with macular thinning”). (Middle) SD-OCT map of right eye of Patient 14 (SS sickle cell patient, best-corrected visual acuity 20/20) showing no macular focal thinning and therefore grouped to Group B (“without macular thinning”). (Right) SD-OCT map of right eye of an age-similar normal control showing normal retinal thicknesses.

Macular Microperimetry Testing

Microperimetry was performed using the OPKO Spectral OCT/SLO system. All tests were performed after dilation of the pupil with 1% tropicamide and 2.5% phenylephrine. The nontested eye was patched during testing. The subjects underwent a practice test and were instructed to fixate on a red square and to depress a button as soon as they saw the stimulus light. The Polar 3 testing grid was used for all subjects. Polar 3 is a standardized grid composed of 28 points arranged in 3 concentric circles (2.3 degrees for inner, 6.6 degrees for middle, and 11 degrees for outer circle in diameter) within the central macula. The inner circle is composed of 4 points, whereas the middle and outer circles are each composed of 12 points. Parameters included a Goldmann III–size stimulus (area of 4 mm ² , diameter of 26 minutes of arc or 0.4 degrees), a 200-millisecond duration of stimulus presentation, and a 4-2 test strategy. The 4-2 test strategy is the standard method in automated static perimetry for estimating threshold. It uses a staircase with an initial crossing of threshold in 4-dB increments and a final crossing in 2-dB increments.

The system automatically tracks fundus localization according to retinal vessel alignment to ensure accurate stimuli placement. If focal macular thinning was found on SD-OCT, a customized testing pattern was created and projected to the area of focal thinning in addition to the standard Polar 3 pattern. After retinal thickness and microperimetry maps were generated, the images were aligned using native software so that the retinal thickness at each microperimetry test location could be obtained.

Normative sensitivity values and intrasession repeatability were measured in 32 normal subjects with the same instrument for comparison with our study cohort. In summary, the sensitivity values averaged across 2 tests for inner, middle, and outer circles ranged from 14.3 to 18.8 dB (median value of 16.9 dB), 13.8 to 18.3 dB (median value of 17.2 dB), and 11.3 to 18.3 dB (median value of 16.6 dB), respectively. The mean intrasession sensitivity variability between 2 tests was small (0.13 dB). Based on these results, retinal sensitivity of 12 dB or lower was considered abnormal for our particular instrument.

Data Analysis

For data analysis purposes, Snellen chart visual acuity was converted to logarithm of minimal angle of resolution (logMAR) visual acuity. Eyes with focal macular thinning were grouped into Group A and those without thinning were grouped into Group B. Two-tailed, unpaired t test was performed between Group A, Group B, and controls using Microsoft Excel software. A P value less than .05 generated by the program was considered statistically significant.