Purpose
To describe the spectrum of retinal capillary ischemia, including superficial and deep capillary ischemia, as identified with spectral-domain optical coherence tomography (SD OCT), that occurs in retinal arterial occlusive disease.
Design
Retrospective observational case series.
Methods
Clinical charts, color fundus photography, red-free fundus photography, fluorescein angiography, near-infrared reflectance, and SD OCT imaging in 40 eyes of 35 patients with retinal arterial occlusive disease were studied in both the acute and chronic phases in multicenter clinical practices. SD OCT imaging analysis was employed to characterize the presence of superficial and deep capillary ischemia in each eye.
Results
Of the 40 eyes, 15 eyes had central retinal artery occlusion (CRAO), 22 eyes had branch retinal artery occlusion (BRAO), and 3 eyes had cilioretinal artery occlusion. During the acute phase, SD OCT showed the following 3 distinct patterns, related to retinal ischemia occurring at varying levels within the retina: (1) thickening and hyperreflectivity of the inner retinal layers, including the nerve fiber and ganglion cell layers owing to ischemia of the superficial capillary plexus; (2) a hyperreflective band at the level of the inner nuclear layer, termed “paracentral acute middle maculopathy,” representing ischemia of the intermediate and deep retinal capillary plexuses (deep capillary ischemia); and (3) diffuse thickening and hyperreflectivity of both the inner and middle retinal layers, which represented both superficial and deep capillary ischemia. Of all eyes, 31 (78%) had both superficial and deep lesions. The remaining 9 eyes (22%) had isolated deep capillary ischemia producing paracentral acute middle maculopathy with sparing of the superficial capillary plexus and a normal fluorescein angiographic appearance. As the lesions evolved into the chronic phase over the ensuing 3 months, the resultant thinning and atrophy reflected the retinal layers affected during the acute phase.
Conclusion
SD OCT imaging reveals the spectrum of capillary ischemia in retinal artery occlusive disease showing variable involvement of the superficial and intermediate/deep capillary plexuses. Isolated deep capillary ischemia manifested as paracentral acute middle maculopathy on SD OCT and may be seen in some eyes with retinal arterial circulation compromise despite complete absence of perfusion abnormalities on fluorescein angiography.
With the advent of spectral-domain optical coherence tomography (SD OCT) imaging, retinal ischemia can be more precisely localized to the superficial and/or intermediate and deep retinal capillary plexuses. The superficial capillary plexus resides in the ganglion cell layer. Superficial capillary ischemia has been well defined in the literature and usually presents clinically as a fluffy “cotton-wool spot” in the acute phase and as a “retinal depression sign” in the chronic phase. The intermediate and deep capillary plexuses reside at the inner and outer border zone of the inner nuclear layer (INL), respectively. Ischemia of these plexuses, deep capillary ischemia, analogous to a deep “cotton-wool spot,” presents as a deeper gray-white lesion with defined edges in the acute phase and evolves into subtle darkening of the retina in the chronic phase.
Fluorescein angiography (FA) has traditionally been the gold standard for evaluating retinal vascular circulation; however, standard FA cannot visualize the intermediate and deep capillary plexuses and therefore may fail to identify deep capillary ischemia. With SD OCT imaging, however, deep capillary ischemia in the acute phase can be recognized as a characteristic hyperreflective lesion at the level of the inner nuclear layer, referred to as paracentral acute middle maculopathy. Paracentral acute middle maculopathy has been described in association with acute macular neuroretinopathy, diabetic retinopathy, retinal vein occlusion, and various retinal vascular disorders. The purpose of this study is to characterize the spectrum of SD OCT findings of eyes with retinal artery occlusion and its association with paracentral acute middle maculopathy.
Methods
This study was approved by the various Institutional Review Boards affiliated with each author, and adhered to the tenets of the Declaration of Helsinki and was conducted in accordance with regulations set forth by the Health Insurance Portability and Accountability Act.
This was a retrospective, nonconsecutive, observational case series, which included patients with the diagnosis of retinal artery occlusion based on clinical findings and ancillary testing, including ophthalmoscopic evidence of retinal whitening, delayed arterial filling with FA, and evidence of macular ischemia with SD OCT imaging. All patients underwent comprehensive ophthalmic assessment, including Snellen visual acuity, slit-lamp biomicroscopy, indirect ophthalmoscopy, color and red-free fundus photography, FA, and SD OCT analysis with simultaneous near-infrared reflectance (NIR) imaging at a central wavelength of 820 nm (Spectralis, Heidelberg, Germany or Cirrus HD-OCT; Carl Zeiss Meditec, Inc, Dublin, California, USA). Retinal arterial occlusive disease was categorized into central retinal artery occlusion (CRAO), branch retinal artery occlusion (BRAO), and cilioretinal artery occlusion based on color fundus photography and FA. The acute hyperreflective lesions on SD OCT imaging were classified into superficial or deep capillary ischemia according to the location and extent of involvement of the retinal layers. The proportion of eyes that displayed both superficial and deep capillary ischemia vs isolated superficial or deep capillary ischemia was analyzed. The lesions on SD OCT imaging were correlated with the clinical appearance and FA.
Statistical analysis of the final visual acuities of patients with CRAO and BRAO were performed using SPSS Software Version 22.0 (IBM Corporation, Armonk, New York, USA), converting Snellen visual acuities to logarithm of the minimal angle of resolution (logMAR). The difference in the mean final visual acuity was calculated using independent sample t test, taking a value of less than .05 as statistically significant.
Results
A total of 40 eyes of 35 patients (15 male and 20 female) with retinal arterial occlusive disease, with a mean age of 61 ± 17.8 years (range 17–91 years), were included in the study. The mean duration of follow-up was 20 ± 26.2 months (range 0–120 months). Of the 40 eyes, 15 eyes had CRAO, 22 eyes had BRAO, and 3 eyes had cilioretinal artery occlusion. Of all 40 eyes, 6 eyes with CRAO and 2 eyes with BRAO had preceding or concurrent central retinal vein occlusion (CRVO). Twenty-four out of 35 (70%) patients had preexisting systemic vascular disease, of which 17 out of 24 (70%) of these could be attributed to hypertension. A summary table showing the demographics and systemic diseases of all patients is included ( Table ).
Patient No | Sex | Age | Eye | BCVA (Onset) | BCVA (Final) | Follow-up (Mo) | Diagnosis | SCI/DCI | Other Eye Diseases | Systemic Diseases |
---|---|---|---|---|---|---|---|---|---|---|
1 | M | 64 | OS | CF | CF | 12 | CRAO | SCI+DCI | HTN | |
2 | F | 81 | OS | CF | CF | 18 | CRAO | SCI+DCI | HTN | |
3 | F | 57 | OS | 20/150 | 20/50 | 6 | CRAO | SCI+DCI | CSC | Non Hodgkin lymphoma, DM, HTN, renal failure, aortic atheroscleros, sleep apnea, emphysema, anemia, DVT, sepsis |
4 | M | 91 | OD | CF | NLP | 120 | CRAO | SCI+DCI | HTN | |
5 | M | 63 | OD | 20/300 | 20/300 | 8 | CRAO | SCI+DCI | ||
6 | F | 57 | OS | 20/400 | 20/400 | 9 | CRAO | SCI+DCI | ||
7 | F | 64 | OD | 20/200 | 20/200 | 7 | Cilioretinal AO | SCI+DCI | HTN | |
8 | F | 80 | OS | 20/80 | 20/40 | 24 | BRAO | SCI+DCI | HTN, anemia | |
OD | 20/50 | 20/30 | 24 | BRAO | SCI+DCI | |||||
9 | F | 49 | OD | 20/20 | 20/30 | 44 | BRAO | SCI+DCI | SLE ×20 years | |
10 | F | 67 | OD | 20/50 | 20/50 | 4 | BRAO | SCI+DCI | ||
11 | F | 33 | OD | 20/20 | 20/30 | 46 | BRAO | SCI+DCI | ||
12 | F | 49 | OD | 20/40 | 0 | BRAO | SCI+DCI | HTN | ||
13 | M | 91 | OS | 20/20 | 0 | BRAO | SCI+DCI | AMD | HTN, s/p aortic stent and bypass surgery | |
14 | F | 55 | OS | CF | 0 | CRAO | Only DCI | Left carotid dissection, left parietal lobe embolic strokes | ||
15 | F | 67 | OS | CF | CF | 3 | BRAO | SCI+DCI | DM, RHD | |
16 | M | 82 | OD | 20/80 | 20/60 | 4 | BRAO | SCI+DCI | POAG | HTN, renal failure, stroke |
17 | F | 80 | OS | CF | CF | 49 | CRVO+CRAO | SCI+DCI | Ocular lymphoma, radiotherapy | Multiple myeloma, non-Hodgkin lymphoma |
OD | 20/25 | 20/25 | 49 | BRAO | Only DCI | |||||
18 | F | 70 | OD | 20/150 | 20/40 | 50 | BRAO | SCI+DCI | NPDR, macroaneurysm | HTN, DM, hypercholesterolemia, arrhythmia/left bundle branch block, mild erythrocytosis and thrombocytosis ×6 years |
19 | M | 17 | OD | 20/15 | 0 | BRAO | SCI+DCI | Prepapillary vascular loop | ||
20 | F | 59 | OD | 20/20 | 0 | BRAO | Only DCI | Aortic calcifications | ||
21 | F | 87 | OS | 20/50 | 5/200 | 9 | CRVO+CRAO | SCI+DCI | Rubeosis | |
22 | M | 80 | OS | CF | CF | 2 | CRVO+CRAO | SCI+DCI | HTN, Parkinson disease | |
23 | F | 45 | OD | HM | HM | 25 | CRVO+CRAO | SCI+DCI | Vasculitis | VSD |
24 | M | 77 | OD | HM | HM | 27 | CRVO+CRAO | SCI+DCI | HTN | |
25 | F | 81 | OS | 4/200 | 4/200 | 8 | CRVO+BRAO | SCI+DCI | ||
26 | M | 63 | OD | 20/25 | 20/25 | 1 | CRVO+BRAO | SCI+DCI | POAG | Multiple myeloma, neuropathy |
27 | F | 55 | OS | 20/200 | 20/100 | 11 | CRVO+CRAO | SCI+DCI | ||
28 | M | 42 | OD | 20/20 | 20/20 | 62 | BRAO | SCI+DCI | NPDR | DM |
OS | 20/20 | 20/20 | 62 | BRAO | SCI+DCI | NPDR | ||||
29 | M | 39 | OD | 20/20 | 20/20 | 11 | BRAO | SCI+DCI | Vasculitis | |
OS | 20/20 | 20/20 | 11 | BRAO | SCI+DCI | Vasculitis | ||||
30 | F | 26 | OD | 20/25 | 20/25 | 5 | BRAO | Only DCI | Factor V Leiden deficiency, hemophilia C, gastric bypass surgery | |
31 | F | 51 | OD | 20/20 | 20/20 | 4 | BRAO | Only DCI | SLE, MI | |
20/200 | 20/200 | 6 | BRAO | SCI+DCI | RD, s/p PPV | |||||
32 | M | 55 | OS | 20/20 | 0 | Cilioretinal AO | Only DCI | HTN | ||
33 | M | 67 | OS | CF | CF | 80 | Cilioretinal AO | Only DCI | Macular hole | HTN, hyperlipidemia |
34 | M | 58 | OD | 20/400 | 20/400 | 1 | CRAO | Only DCI | ||
35 | M | 44 | OD | CF | 0 | CRAO | Only DCI |
The acute phase of retinal arterial occlusive disease was studied in 35 eyes since 5 eyes presented at baseline in the chronic phase, while the chronic phase was studied in 38 eyes owing to lack of follow-up in 2 eyes. With SD OCT imaging, the acute phase showed 3 types of patterns, depending on the level of involvement of the retinal layers: (1) thickening and hyperreflectivity of the inner retinal layers, including the nerve fiber and ganglion cell layers, owing to ischemia of the superficial capillary plexus; (2) a hyperreflective band at the level of the inner nuclear layer, also termed paracentral acute middle maculopathy, that represented ischemia of the intermediate and deep retinal capillary plexuses; and (3) diffuse thickening and hyperreflectivity of both the inner and middle retinal layers, which represented ischemia of the superficial, intermediate, and deep capillary plexuses. These lesions could be found at varying locations throughout the posterior pole but were always paracentral and were also identified at varying phases throughout the course of the disease. The chronic phase was seen to occur between 1 and 3 months from the acute baseline presentation and showed resultant thinning and atrophy of the retinal layers, corresponding to the acute lesions, when present. Of note, intermediate and deep retinal capillary ischemia never occurred exclusive of the other; and therefore we refer to deep capillary ischemia to include both levels of involvement.
Of all 40 eyes, 31 (78%) demonstrated evidence of superficial and deep capillary ischemia in the same eye as both independent and contiguous lesions. None of the eyes showed only superficial capillary ischemia in the absence of deep capillary ischemia. In 7 eyes with contiguous lesions, the core of the lesion demonstrated superficial and deep capillary ischemia while the border zone of retinal whitening in the perifoveal region showed deep capillary ischemia. Nine eyes (22%) showed only deep capillary ischemia, manifested as paracentral acute middle maculopathy on SD OCT. Isolated deep capillary ischemia could be seen in CRAO (n = 4), BRAO (n = 4), and cilioretinal artery occlusion (n = 1) cases. FA failed to show any identifiable perfusion abnormality in 8 out of these 9 cases. The lesions with superficial capillary ischemia corresponded to areas of fluffy inner retinal whitening, similar to a cotton-wool spot, and appeared hyporeflective with NIR imaging, although not as prominently dark as the deeper lesions. The paracentral acute middle maculopathy lesions clinically corresponded to areas of milder and deeper retinal whitening and were more prominently hyporeflective with NIR. In cases of isolated paracentral acute middle maculopathy and deep capillary ischemia, there was no evidence of inner retinal involvement at any time by virtue of the absence of inner retinal hyperreflectivity and inner retinal thinning. In all 15 eyes with CRAO, SD OCT showed varying degrees of hyperreflectivity at the fovea corresponding to a cherry-red spot and attributable to increased transmission of light relative to the adjacent opacified perifoveal retina.
The mean final visual acuity was counting fingers in eyes with CRAO and 20/40 in eyes with BRAO. In eyes with CRAO, the mean final visual acuity was equally poor in eyes with isolated deep capillary ischemia vs eyes with both superficial and deep capillary ischemia ( P = .77). Similarly, in eyes with BRAO, there was no difference in the mean final visual acuity in eyes with isolated deep capillary ischemia compared to eyes with both superficial and deep capillary ischemia ( P = .45).
Here, we describe some representative cases to showcase the spectrum of SD OCT findings in retinal arterial occlusive disease.
Case 1
An 80-year-old woman (Patient 8) with systemic hypertension and anemia presented with sudden vision loss and visual acuity of 20/80 in the left eye. Clinical examination showed BRAO with retinal whitening along the superotemporal arcade and FA showed delayed perfusion of the superotemporal branch retinal artery. Within the area of retinal whitening during the acute phase, SD OCT showed thickening and hyperreflectivity of both the inner and middle retinal layers, with sparing of the outer retinal architecture. In the chronic phase 2 years later, SD OCT through the same area showed thinning and atrophy of the inner and middle retinal layers. Immediately adjacent to this area at the edge of retinal whitening toward the fovea, SD OCT showed hyperreflectivity of the middle layers at the level of the INL consistent with paracentral acute middle maculopathy in the acute phase. In the chronic phase, SD OCT through this area demonstrated thinning of only the INL. FA in the chronic phase failed to reveal any evidence of ischemia or nonperfusion. In summary, this case of BRAO showed evidence of both superficial and deep capillary ischemia in 1 region and only deep capillary ischemia in a separate region of the macula. This case also clearly demonstrated that FA would fail to identify a BRAO in the chronic phase, whereas SD OCT may be more helpful in demonstrating evidence of previous retinal ischemia ( Figure 1 ).
Case 2
A 91-year-old man (Patient 13) with systemic hypertension and cardiovascular disease presented with sudden painless superior visual field loss in the left eye. Visual acuity was 20/20 at baseline despite the presence of nonexudative age-related macular degeneration. Clinical examination in the acute phase showed scattered areas of retinal whitening along the inferior branch retinal artery while FA demonstrated delayed perfusion in the inferior branch retinal artery consistent with an inferior BRAO. SD OCT imaging showed thickening and hyperreflectivity of the inner retinal layers adjacent to the optic nerve, consistent with superficial capillary ischemia, and a cotton-wool spot. Thickening and hyperreflectivity of both the inner and middle retinal layers at the level of the INL were noted temporally, indicating superficial and deep capillary ischemia ( Figure 2 ).