Optical coherence tomography angiography in choroidal haemangioma: small case series.

Abstract

In the present study, we described the vascular features of circumscribed choroidal haemangiomas (CHs) by means of optical coherence tomography angiography (OCT-A) in a case series of three patients. Optical coherence tomography angiography (OCT-A) is a recently developed technique, which allows structural imaging of retinal circulation and in vivo analysis of microvascular alterations without intravenous dye injection (Jia et al. 2014, 2015; De Carlo et al. 2015; Ishibazawa et al. 2015). In this case series, OCT-A was performed using spectral domain optical coherence tomography (SD-OCT) RTVue XR AVANTI (Optovue Inc, Fremont, CA, USA), a 70 kHz SD-OCT system with a centre wavelength of 849 nm. This device employs a new analytical algorithm, ‘split-spectrum amplitude-decorrelation angiography with optical coherence tomography’ (SSADA), to obtain anatomical and functional information of retinal circulation, imaging both the microvascular network and the blood flow in the selected area (Jia et al. 2012). To allow a better visualization, we chose the deep choroid capillary (CC) slab on OCT-A. High-quality OCT angiograms of the intralesional vascular pattern have been obtained by segmenting a CC slab (from 30 to 70 μm) below Bruch's membrane in the angiovue software. Although the 6 × 6 or the 8 × 8 mm scans are recommended for studying large areas of the retina, a 3 × 3 mm scan was preferred to better visualize the intralesional vasculature (Fig. 1A–D). A large-field en-face OCT angiograms (~3 × 8.5 mm) were then produced by stitching together three 3 × 3 mm scans in case of large CHs (Fig. 1E). Patients were 37, 48 and 56 years old, respectively. Ophthalmic features of the three CHs revealed a mean largest diameter of 6.4 mm and a mean thickness of 3.1 mm. The tumours were located within the temporal vascular arcades in two eyes and inferiorly to the optic disc in one patient. The en-face OCT-A was able to reveal intralesional vascular pattern in all cases. More centrally, tumours were surrounded by a richly branched and convoluted capillary pattern with a high flow signal, while more peripherally, they were wrapped by rarefied vessels. A high flow capillary network was mainly shown in all lesions, composed by large blood-filled vascular channels and separated by intervascular septa where flow was not detected. At the tumour periphery, numerous radially projecting vessel ends were observed. In one case, the inner slab segmentation also showed extensive coalescent spaces hyporeflective on OCT-A within the inner retinal layers and corresponding to intraretinal cystic changes overlying the lesion. The OCT-A ability to non-invasively visualize pathological area offered advantages over dye-based angiography methods. Moreover, data are three-dimensional and depth-resolved, without leakage obscuring the vascular structures. Currently, fluorescein angiography (FA) and indocyanine green angiography (ICGA) are considered the gold standard for defining CHs features (Shields et al. 2001). Although this study is not a comparative one, OCT-A visualize both small capillary and large hypereflective vessels inside the tumour separated by absent or little connective tissue more precisely than ICGA in all patients. Our case series had several limitations. It was a small retrospective cross-sectional case series. Selected CHs were located at the posterior pole and easy to scan because of OCT-A limited field of view. Field of view can be expanded, but lowering scan resolution. In our cases series, we used a 3 × 3 scan box that allowed the visualization of a very limited area but a better scanning density in each tumour. Projection artifacts were sometimes observed on the cross-sectional angiograms. The flow signal fade-out in large vessels with very fast blood flow can induce fringe washout of OCT signals. This means that some large vessels in the deep of the tumours could not be visualized using SSADA. Finally, the OCT-A flow signal must be interpreted with care. It is linearly related to velocity over a limited range. If blood flow is extremely slow in pathological conditions, the decorrelation values may be below background noise; in this case, flow would not be detected by OCT-A. Differentiate varying degrees of flow impairment by varying the interscan time could avoid misleading interpretation of the OCT-A images (Choi et al. 2015). In conclusion, OCT-A is a non-invasive imaging tool able to detect blood flow pattern in CHs. Vascular pattern has been shown as clearly as ICGA, and blood flow signal inside the CHs showed a precise and well-delineated capillary-like pattern. With the technological improvement, we believe OCT-A will become an important diagnostic tool also in visualization of deep choroidal lesions.