From the Department of Ophthalmology (H. Chung, D. Kim, J.-G. Kim, J.Y. Lee, J.Y. Lim, Y.H. Yoon) and Department of Surgery (S.-H. Ahn), Asan Medical Center, University of Ulsan, College of Medicine, Seoul, Korea.
Supported by a grant from the Korea Heathcare Technology R&D Project, Ministry of Health, Welfare & Family Affairs, Republic of Korea (#A080557).
The authors have no financial or proprietary interest in the materials presented herein.
Address correspondence to Young Hee Yoon, MD, Department of Ophthalmology, Asan Medical Center, 388-1 Pungnab-dong, Songpa-gu, Seoul, 138-736, Korea.
In 1978, Kaiser–Kupfer and Lippman reported that tamoxifen can cause keratopathy and retinopathy.1 Recently, Gualino et al.2 and Martine et al.3 used time-domain optical coherence tomography (TD-OCT) to characterize three cases of tamoxifen retinopathy. Because tamoxifen mostly affects the macular region, high-quality imaging modalities that reveal fine macular pathology should be used. OCT generates cross-sectional images by measuring the echo time delay and the magnitude of backscattered or back-reflected light.4 The Stratus OCT system (Carl Zeiss Meditec, Dublin, CA) has an axial resolution of 10 μm. 3D-OCT (using spectral or Fourier domain detection) has enabled better delineation of intraretinal layers by combining ultrahigh resolution (~3.5 μm) and high data acquisition speeds (87,000 A-scans in ~4 s).
Here, we report on the presence of pathogenic macula in 6 patients who had taken low cumulative doses (4.2 to 9.6 g) of tamoxifen.
Design and Methods
From January 2007 to January 2008, 49 breast cancer patients undergoing tamoxifen treatment (20 mg per day) were examined at the eye clinic of the Asan Medical Center (Seoul, Korea). All patients either had subjective symptoms or were examined during the course of routine checkups. Eye examinations were conducted 4 to 60 months after commencement of tamoxifen treatment. For all patients, we recorded a detailed history and performed ophthalmic tests including dilated fundus examinations, TD-OCT (OCT 3 instrument; Stratus, Carl Zeiss Meditec, Dublin, CA), and 3D-OCT (3D-OCT 1000 instrument; Topcon, Tokyo, Japan). Clinical examinations included fluorescein angiography (FA), electroretinography (ERG), and visual field tests in selected patients.
Eight patients showed abnormal 3D-OCT change, and the changes in 2 cases were attributed to age-related macular degeneration. In this report, we discuss the 3D-OCT findings of 6 patients whose subfoveal structural changes appeared to be caused by low cumulative doses of tamoxifen.
Six patients had abnormal 3D-OCT results. Four of these patients complained of mild decreases in visual acuity and two had no symptoms. All 6 patients had taken cumulative doses of 4.2 to 9.6 g of tamoxifen over the previous 7 to 16 months.
3D-OCT indicated foveal cystoid spaces in 10 eyes and foveolar cystoid spaces in 6 eyes (Table 1). We identified cystoid spaces in the foveal regions of eight eyes by TD-OCT. TD-OCT did not indicate detailed microstructural changes in the affected retinas nor did the technique yield the precise locations of cystoid spaces.
Table 1: Clinical, Fundus, FA, TD-OCT, and 3D-OCT Data on Patients Receiving Tamoxifen
We used longitudinal cross-sections provided by 3D-OCT to determine the precise location of each cystoid space (Table 1, Figure 1). These indicated no evidence of cystoid macular edema or macular thickening. Instead, 2 eyes showed macular thinning (Patients 1 and 2). Our results differ from those of previous reports that described tamoxifen-induced retinopathy manifesting principally as cystoid macular edema.1,5,6 Our results indicate some degree of atrophy of retinal tissue in the foveal center, as also indicated by the disruption of photoreceptor transition lines.
Figure 1. TD-OCT and 3D-OCT of Patients 1 (a) and 2 (b). Horizontal and Vertical Scans, Using the Stratus OCT, were Taken at Sites that were Presumptively Chosen. These Failed to Show Definite Abnormalities in the Foveae of the Left Eyes of Patients 1 and 2. Consecutive Orthogonal Cross-Sectional Images from 3D-OCT Enabled Detection of Intraretinal Cysts and Focal Disruptions of the Photoreceptor Transition Lines in Both Patients.
In our study, patients with foveolar intraretinal cysts were more likely to have taken larger total dosages of tamoxifen than had patients with extrafoveolar cysts (Mann–Whitney U, P = .010). Yellow-white refractile deposits, which appeared as focal perifoveal and foveolar hyperreflective spots in the inner layer on 3D-OCT, were present in the foveae of Patient 1. 6 of the eyes with abnormal 3D-OCT findings showed only decreased foveal reflexes in dilated fundus examinations. Two eyes with tiny cysts on 3D-OCT showed no fundic abnormalities. In 8 eyes, FA showed foveal hyperfluorescence of varying degrees, with no leakage or pooling throughout the entire sequence. This early hyperfluorescence was the only abnormal finding seen in FA and was related to the foveolar location of intraretinal cysts (Fisher’s exact test, P = .033). The extent of cyst or photoreceptor transition line defect was not related to this FA finding. Patients with normal FA results had better visual acuity than had patients with hyperfluorescence on FA (Mann–Whitney U, P = .044). No patient yielded abnormal results ERG, or visual field tests.
Because only a limited set of B-scans (tomograms) were taken at sites presumptively chosen during initial examination, TD-OCT techniques such as Stratus OCT cannot provide detailed evaluations of the early stages of maculopathy and cannot effectively monitor disease progression. On the other hand, spectral-domain, ultra-high resolution 3D-OCT (87,000 A scans in ~4 s) provides comprehensive coverage of the retina with a reduced number of sampling errors.4 Consecutive orthogonal cross-sectional imaging allows comprehensive observation of the microstructure and extent of cystoid spaces. The combination of ultrahigh resolution and high-speed data acquisition provided by 3D-OCT provides clear delineation of intraretinal layers.
This report is the first to describe maculopathy that was induced by very low cumulative doses of tamoxifen. Each patient had a cumulative tamoxifen dose of less than 10 g over a period of less than 16 months. When evaluated with TD-OCT, we observed intraretinal cysts in 8 of 12 eyes. However, the degree of photoreceptor layer disruption and the exact locations and extents of cysts were not visible using TD-OCT.
Clear visualization of all major intraretinal layers and the ability to detect changes in retinal morphology (especially in the photoreceptor transition line) early during tamoxifen therapy can have a significant impact on the diagnosis and treatment of tamoxifen-induced maculopathy. Patient 1 developed maculopathy very early in treatment and suffered from significant visual impairment. Compared with patients who had extrafoveolar cysts, patients with foveolar cysts were likely to have taken larger total dosages of tamoxifen. However, the extent of cysts (entire retina, or inner or outer layers), the presence of photoreceptor transition line defects, and central retinal thickness, were not related to total tamoxifen dosage. Eyes seemed to retain good visual acuity when foveal cystoid spaces were extra-foveolar and did not involve the entire retinal layer, or when photoreceptor transition lines were intact. However, these associations were not statistically significant. Cystoid spaces located at the extrafoveolae do not involve the entire retinal layer. However, 4 of 6 cysts that we observed at the foveolae involved all layers of the retina.
Tamoxifen retinopathy is rare and the need to discontinue tamoxifen in asymptomatic patients with limited ocular changes has not been established. However, we recommend 3D-OCT screening for all patients treated with tamoxifen. This includes patients with no subjective symptoms, with no definite fundic abnormalities, and with no clear abnormalities on TD-OCT. Screening with 3D-OCT is required for the early detection and prevention of tamoxifen-induced macular degeneration. This approach will be particularly useful when patients on tamoxifen complain of visual changes but show no clear abnormalities upon routine ophthalmologic testing. Under such circumstances, we recommend close monitoring of possible visual disease progression by 3D-OCT and measurement of changes in visual acuity.
The clear delineation of intraretinal layers provided by 3D-OCT will improve our understanding of the pathogenesis of tamoxifen-induced maculopathy. The appearance of tiny intraretinal cysts or disruption of the photoreceptor transition line (both demonstrating tamoxifen toxicity) indicate that drug treatment should be discontinued or that close monitoring of visual disease progression is needed. All physicians should be aware of the potential for ocular toxicity associated with tamoxifen, even when the cumulative dose is very low. Doctors should encourage baseline ophthalmologic evaluations and follow-up monitoring of any ocular complaints that arise during tamoxifen therapy.
- Kaiser-Kupfer MI, Lippman ME and . Tamoxifen retinopathy. Cancer Treat Rep. 1978;62:315–320.
- Gualino V, Cohen SY, Delyfer M-N, Sahel J-A, Gaudric A. Optical coherence tomography findings in tamoxifen retinopathy. Am J Ophthalmol. 2005;140:757–758. doi:10.1016/j.ajo.2005.04.042 [CrossRef]
- Martine M-F, Joël G, Maddalena Q-EM. Optical coherence tomography in tamoxifen retinopathy. Breast Cancer Research and Treatment. 2006;99:117–118. doi:10.1007/s10549-006-9187-y [CrossRef]
- Drexler W, Fujimoto JG. State-of-the-art retinal optical coherence tomography. Progress Retinal Eye Res. 2008;27(1):45–88. doi:10.1016/j.preteyeres.2007.07.005 [CrossRef]
- Kaiser-Kupfer MI, Kupfer C, Rodrigues M. Tamoxifen retinopathy. A clinicopathologic report. Ophthalmology. 1981;88:89–93.
- Griffiths MFP. Tamoxifen retinopathy at low dosage [letter]. Am J Ophthalmol. 1987;104:185–186.
Clinical, Fundus, FA, TD-OCT, and 3D-OCT Data on Patients Receiving Tamoxifen
|Patient/Eye||Sex/Age||Cumulative Dose (g)||Subjective Symptom||Visual Acuity||Fundus Abnormality||Time Domain OCT||3D-OCT||FA Findings|
|Cystic Space||CFT (μm)||Location, Extent of Intraretinal Cysts||Photoreceptor or Transition Line|
|1 OD||F/53||7.2||+||20/100||yellow crystal||+||178||foveola (sup) entire layer, thin retinal bridge||focal disruption||early foveolar hyperf.|
|1 OS||20/30||yellow crystal||−||156||foveola (sup) entire layer, thin retinal bridge||focal disruption||early foveolar hyperf|
|2 OD||F/43||4.8||+||20/20||FVR||171||no cystic space found||intact||early foveolar hyperf|
|2 OS||20/40||FVR||−||166||fovea (sup) outer layer, transition zone||focal disruption||early foveolar hyperf|
|3 OD||F/48||9.6||+||20/25||FVR||+||187||foveola (sup) inner layer||intact||parafoveolar hyperf.|
|3 OS||20/40||FVR||+||203||foveola (center) entire layer, thin retinal bridge||focal disruption||parafoveolar (temporal) hyperf|
|4 OD||F/58||8.1||+||20/40||FVR||+||198||foveola (center) entire layer, thin retinal bridge||intact||early foveolar hyperf.|
|4 OS||20/40||FVR||+||187||foveola (center) inner layer||focal disruption||early foveolar hyperf.|
|5 OD||F/55||5.4||−||20/30||no||+||183||fovea (sup) inner layer||intact||no abnormality|
|5 OS||20/25||no||+||184||fovea (sup) inner layer||intact||no abnormality|
|6 OD||F/43||4.2||−||20/20||FVR||+||193||fovea inner layer,||intact||no abnormality|
|6 OS||FVR||−||191||no cystic space found||intact||no abnormality|