Foveal cone loss in tamoxifen maculopathy: a case report

Tamoxifen is used in low dose concentrations (20–40 mg per day) as a therapy for breast cancer patients. Clinical ocular imaging has revealed side effects including crystalline deposits in the nerve fiber and inner plexiform layers [1‐5], pseudocystoid changes at the fovea [1, 2, 6‐8], as well as loss of blood flow in the deep capillary plexus [9]. This report describes the high-resolution imaging findings on a tamoxifen patient who complained of a small central scotoma in the left eye while reading. However, subtle loss of cone photoreceptors was also observed in the right eye, which was not visible in standard clinical imaging.

A dilated fundus examination revealed a duller foveal light reflex and a yellowish foveal discoloration in the left eye. However, fundus photography (FP), fundus autofluorescence (FAF), and fluorescein angiography (FA) (Fig. 1A–C) were unremarkable. FP, FAF, and FA for the right eye were also unremarkable (Fig. 1E–G). The FAF images for either eye did not show any areas of abnormal hyper- or hypo-autofluorescence. All other findings including optic discs, vessels, and peripheral retina were unremarkable in both eyes.

Figure 1D shows the spectral spectral-domain optical coherence tomography (SD-OCT) image from the left eye showing focal disruption of cone outer segment (OS) layer at the foveal center with a localized hyper-reflective region located in the cone inner segment (IS) layer. The external limiting membrane (ELM) appears intact. The SD-OCT for the right eye shows a normal macula with contiguous outer retinal layers (Fig. 1H). The inner retinal layers appear to be normal across the retina for both eyes. Both FAF and SD-OCT images were acquired using Spectralis OCT (Heidelberg Engineering Inc., Heidelberg, Germany).

After observing the subtle foveal irregularities in the photoreceptor layer in the left eye from the SD-OCT image, the patient was imaged on a research-grade AO-SLO system. The system hardware and imaging procedure are described in detail in Wells-Gray et al. [10]. Briefly, the AO-SLO acquires 1° × 1° images of the retina at 60 frames per second at an imaging wavelength of 680 nm with a lateral resolution of ~ 2.3 µm. Prior to imaging, a combination of 1% tropicamide and 2.5% phenylephrine was used to dilate the pupil and paralyze accommodation. During imaging, the patient looked at a fixation target displayed on a computer monitor, which corresponded to the area of imaging. Multiple AO-SLO datasets were acquired around the foveal center of both eyes and subsequently postprocessed to remove eye motion. For the AO imaging, the tenets of the Declaration of Helsinki were observed, and the protocol was approved by the Institutional Review Board of The Ohio State University (OSU). Written informed consent was obtained after all procedures were fully explained to the patient and prior to imaging.

The montage of AO-SLO images for the left and right eyes, respectively, are shown (Fig.2A, B). Lesions are more clearly visible in the left eye as compared with the right. Cone densities (using a 50 × 50 µm window) measured at (i) fovea, (ii) 75 µm superior, and (iii) 150 µm superior from the foveal center (just outside of the lesion region) were 26,000, 29,600 and 53,200 cones/mm², respectively, for the affected left eye. The measured values close to the foveal center are much lower than the > 100,000 cones/mm² expected from the literature [11], however the result at 150 µm from the foveal center is in good agreement with the expected value for age-matched healthy subjects [12]. For the right eye, the foveal cone density was 39,200 cones/mm², greater than that measured in the left eye but still well below the expected value for an age-matched control. Cone densities at 75 µm superior and 150 µm superior from the foveal center were 52,000 and 58,400 cones/mm², respectively, in good agreement with published results [11].

Previous SD-OCT studies have shown disruption to the photoreceptor layer in association with tamoxifen use, which is in agreement with the findings presented here here [1, 2, 6‐8, 13, 14]. While inner retina deposits and vascular changes have also been reported [1‐5, 9], these were not observed in our case. Only the loss of foveal cones was revealed in both the symptomatic (left) and asymptomatic (right) eyes using the AO-SLO, and only in the symptomatic eye using a clinical OCT system.

In our patient, all the layers anterior and posterior to the photoreceptor layer appeared to be regular and intact based on the SD-OCT images. Perifoveal telangiectasia was discounted based on the normal fluorescein angiography. While the high-resolution AO-SLO imaging of the lesions did show areas of foveal cone loss, the cone photoreceptor mosaic 150 µm from the foveal center for the left eye and outside the umbo (right eye) were normal, as shown by the cone density measurements.

The subject had previously undergone LASIK surgery, but, unfortunately, their refractive error prior to this surgery was not known. It is well established that myopic eyes have reduced cone density compared with their age-matched emmetropic or hyperopic eyes primarily due to the retinal stretching caused by the axial elongation [15]. Any pre-LASIK myopia would not explain the findings reported here as the cone density changes were limited to the center of the fovea only and normal cone densities were found immediately outside this area.

Solar retinopathy [16] was discounted, as one would expect the retinal changes to be more or less similar in both eyes and for the typical hyper-reflectivity observed in such cases to span multiple retinal layers. Furthermore, FAF and FA for both eyes were unremarkable contrary to what is typically found in solar retinopathy. [17]

To explain the patient’s complaint when reading, lesion size was compared to standard letter size while reading newsprint (Fig. 2A). When compared to a scaled letter “e” representative of text size [18] while reading at a distance of 40 cm, it appears the central lesion in the left eye is large enough to obscure single letters while reading, potentially explaining her central scotoma.

The pathophysiology of tamoxifen maculopathy is yet to be understood, but it is thought to be related to dysfunction and degeneration of Müller cells [19]. Given the isolated loss of foveal cone photoreceptors with the absence of previously reported inner retinal and vascular changes, our results may indicate the earliest changes associated with tamoxifen retinopathy. Furthermore, SD-OCT failed to show any structural changes in the asymptomatic eye (right eye) while the AO-SLO was able to detect smaller focal areas of cone loss, demonstrating the ability to detect structural changes before manifestation of functional loss noticed by the patient.