Incidence and risk factors of massive subretinal hemorrhage in retinal angiomatous proliferation

Abstract

Objective: To evaluate the incidence and associated risk factors of massive subretinal hemorrhage (SRH) in patients with retinal angiomatous proliferation (RAP).

Methods: A total of 187 eyes of 135 treatment-naıve patients diagnosed with RAP were evaluated retrospectively. Clinical records including the time between the initial visit, last anti-vascular endothelial growth factor (VEGF) treatment, last stable examination, and the date of massive SRH were reviewed. Imaging findings including indocyanine green angiography (ICGA) and optical coherence tomography (OCT) were analyzed.

Results: Massive SRH developed in 18 eyes (9.6%) a median of 20 months after the initial presentation. Kaplan-Meier survival analysis revealed that the incidence (2.8, 5.8, 13.1, and 21.0% after 1,2,5 and 10 years, respectively) continuously increased. Among 14 eyes with discernable vascular anastomosis on baseline ICGA, 13 (92.8%) showed retinal arteriole involvement. On spectral-domain OCT imaging of the last visit prior to the massive SRH, a layered lamellar tissue complex was noted under the retinal pigment epithelium in 9 of 13 eyes, which was significantly associated with massive SRH[hazard ratio(HR),5.883;P = .010]. The average time between the last stable examination/last injection and the massive SRH was 2 and 5 months, respectively. The patients were treated with anti-VEGF, gas and recombinant tissue plasminogen activator injection; however, all except one eye had visual acuity worse than 20/1000 at the final visit.

Conclusions: Massive SRH can occur in RAP in the course of anti-VEGF treatment, resulting in severe vision loss. A proactive dosing regimen may be more appropriate for these RAP eyes.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Fig 1. Cumulative incidence of massive subretinal hemorrhage (SRH) in retinal angiomatous proliferation (RAP) according to a Kaplan—Meier survival analysis.

The horizontal axis shows the time from baseline (treatment initiation with for RAP), and the vertical axis shows the proportion of eyes that developed SRH. The incidence rates of massive SRH were 2.8%, 5.8%, 13.1%, and 21.0% of eyes at 1, 2, 5, and 10 years, respectively, after the initial treatment.

Fig 2. Massive subretinal hemorrhage (SRH) in patient 14, a 69-year-old female with retinal angiomatous proliferation (RAP).

(A) At baseline (July 2007), the color fundus photograph showed multiple drusen and retinal hemorrhages inferior to the fovea. (B) Fluorescein angiography (FA) showed intraretinal leakage and blocked fluorescence due to retinal hemorrhage. Inset: Vertical scan through the fovea on Stratus optical coherence tomography (OCT) demonstrated cystic space in the neurosensory retina and pigment epithelium detachment (PED). (C) Mid-to-late phase indocyanine green angiography (ICGA) revealed a small hot-spot lesion, in which an anastomosis between the retinal arteriole (a) and venule (v) was detected on mid-phase ICGA (inset). Best-corrected visual acuity (BCVA) of the left eye was 20/40, and the patient received a combination treatment of photodynamic therapy and bevacizumab injection. After the initial treatment, six ranibizumab and eight bevacizumab injections were performed during the follow-up period; however, the BCVA decreased to 20/200 at 97 months after the initial treatment. (D) Sectional spectral domain (SD)-OCT of the fovea taken at that visit showed well-organized layered lamellar tissue within fibrovascular PED (arrows) as well as atrophic retinal pigment epithelium (RPE) and increased signal transmission in the choroid. (E) Near-infrared images confirmed the development of geographic atrophy during the follow-up period. (F) One month later, massive SRH extending to the major vascular arcade developed. (G) SD-OCT showed subretinal and sub-RPE hemorrhage involving the fovea. The patient received a single bevacizumab injection with gas/rt-PA injection, and 6 months later, BCVA decreased to hand motion. (H) SD-OCT showed the contraction of layered lamellar tissue, and as a result, hyporeflective space was noted, separating the neovascular tissue from the underlying choroid and Bruch’s membrane.

Fig 3. Massive subretinal haemorrhage (SRH) in Patient 11, a 71-year-old woman with retinal angiomatous proliferation (RAP).

(A) At baseline, the color fundus photograph showed multiple drusen, retinal hemorrhage, and a neovascular membrane. (B) Fluorescein angiography revealed dye leakage from the neovascular lesion and cystoid changes at the fovea. (C) Mid-phase indocyanine green angiography with magnification (inset). A neovascular complex with retinochoroidal anastomosis composed of a perfusing retinal arteriole (a) and two draining venules (v) was noted. (D) Massive SRH extending past the equator occurred at 21 months after the baseline, and 11 months after the last ranibizumab injection. Best-corrected visual acuity of the right eye decreased from 20/100 to 20/1000. (E) Spectral-domain optical coherence tomography images corresponding to the white lines on color fundus photography (inset of D) revealed intraretinal edema with layered lamellar tissue within fibrovascular tissue under the retinal pigment epithelium.

Fig 4. Development of a retinal pigment epithelium (RPE) tear after massive subretinal hemorrhage (SRH) in Patient 3, a 76-year-old woman with retinal angiomatous proliferation (RAP).

(A) Mid-to-late phase indocyanine green angiography (ICGA) at baseline revealed a small hot-spot lesion. Inset: Early phase ICGA showing a vascular anastomosis involving retinal arteriole (a). (B) Spectral-domain optical coherence tomography through the lesion demonstrated intraretinal hyperreflective lesion with pigment epithelium detachment. (C) Color fundus photograph at baseline showing retinal hemorrhage and exudate deposition around the hot-spot lesion on ICGA. (D) Massive SRH developed 12 months after the baseline. The patient received a single bevacizumab injection with gas/rt-PA injection. (E) Then, 2 months later, an RPE tear developed (arrowheads) and best-corrected visual acuity of the left eye decreased to counting fingers.

Fig 5. Enlargement of widespread fibrovascular tissue (disease progression) before the development of massive subretinal hemorrhage (SRH) in Patient 4, a 65-year-old woman with retinal angiomatous proliferation (RAP).

(A,B,C) Color fundus photograph, fluorescein angiography (FA) and indocyanine green angiography (ICGA) at baseline (Sept. 2006). (C) Late-phase ICGA revealed a hot spot lesion with intraretinal leakage and pigment epithelial detachment. Inset: Early phase ICGA showing a vascular anastomosis between the retinal arteriole (a) and venule (v). (D,E,F) Color fundus photograph, FA and ICGA taken 33 months after baseline. Eight ranibizumab injections were performed during the follow-up period, and best-corrected visual acuity (BCVA) of the left eye decreased, from 20/100 to 20/160. (F) Late-phase ICGA revealed a large plaque lesion at the macula, indicating the enlarged fibrovascular tissue complex. (G) The patient received another ranibizumab injection 3 months later, and massive bullous SRH extending past the equator developed 3 weeks after the last injection. Then, 1 month later, breakthrough vitreous hemorrhage occurred and a pars plana vitrectomy was performed. (H) At the final visit (78 months after baseline, Mar. 2013), BCVA of the left eye was hand motion due to a diffuse submacular fibrous scar.