Ophthalmic Surgery, Lasers and Imaging Retina

Case Report 

Histopathology of Peripapillary Choroidal Neovascularization

Jose Maria Ruiz-Moreno, MD, PhD; Javier A. Montero, MD, PhD; Laurent Jonet, PhD; Jean-Claude P. Jeanny, PhD; Francine Behar-Cohen, MD, PhD

Abstract

The different therapeutic responses observed among choroidal neovascularization (CNV) of different etiologies, ages, and locations might be related to the presence of varied mediators. Two surgically removed peripapillary CNVs from two different patients were analyzed. One of the patients had received one intravitreous injection of bevacizumab 3 months earlier. CNV was analyzed using conventional histology and immunohistochemistry. Histological analysis showed intense neovascularization and epithelial and glial components. Vascular endothelial growth factor (VEGF) receptors were found in the endothelial cells and the epithelial cells of the CNV. VEGF was expressed in the patient who had not been previously treated with anti-VEGF. The CNV was deeply infiltrated by glial cells and invaded by microglial cells in one case. VEGF and VEGF receptors may be expressed, suggesting that therapies aiming at VEGF may be efficient only for a subtype of CNV and at a certain time point of their evolution.

Abstract

The different therapeutic responses observed among choroidal neovascularization (CNV) of different etiologies, ages, and locations might be related to the presence of varied mediators. Two surgically removed peripapillary CNVs from two different patients were analyzed. One of the patients had received one intravitreous injection of bevacizumab 3 months earlier. CNV was analyzed using conventional histology and immunohistochemistry. Histological analysis showed intense neovascularization and epithelial and glial components. Vascular endothelial growth factor (VEGF) receptors were found in the endothelial cells and the epithelial cells of the CNV. VEGF was expressed in the patient who had not been previously treated with anti-VEGF. The CNV was deeply infiltrated by glial cells and invaded by microglial cells in one case. VEGF and VEGF receptors may be expressed, suggesting that therapies aiming at VEGF may be efficient only for a subtype of CNV and at a certain time point of their evolution.

Histopathology of Peripapillary Choroidal Neovascularization

From the Department of Ophthalmology (JMR-M), Albacete Medical School, Castilla La Mancha University, Albacete, Spain; the Alicante Institute of Ophthalmology (JMR-M, JAM), VISSUM, Vitreo-Retinal Unit, Alicante, Spain; the Department of Ophthalmology (JAM), Pio del Rio Hortega University Hospital, Valladolid, Spain; Centre de Recherche des Cordeliers (LJ, J-CPJ, FB-C), Université Paris Descartes, Paris, France; and the Department of Ophthalmology (FB-C), Université Paris Descartes, Paris, France.

Presented at the 11th Euretina meeting, May 29, 2011, London, United Kingdom.

Supported in part by a grant of the Spanish Ministry of Health, Instituto de Salud Carlos III, Red Temática de Investigación Cooperativa en Salud “Patología ocular del envejecimiento, calidad visual y calidad de vida” (RD07/0062).

The authors have no financial or proprietary interest in the materials presented herein.

Address correspondence to Jose Maria Ruiz-Moreno, MD, PhD, Departamento de Ciencias Médicas, Facultad de Medicina, Avenida de Almansa, 14, 02006 Albacete, Spain. E-mail: josemaria.ruiz@uclm.es

Received: August 04, 2011
Accepted: January 27, 2012
Posted Online: March 15, 2012

Introduction

Peripapillary choroidal neovascularization (CNV) may be associated with varied ocular conditions. Different therapeutic modalities have been reported to treat peripapillary CNV, including intravitreal anti-angiogenic drugs. However, the clinical response to anti-angiogenic treatment may differ among the different cases.

The pathology of submacular CNV associated with age-related macular degeneration (AMD) has been reported in previous series, but reports on the pathology of peripapillary CNV are scarce and have not analyzed the presence of the vascular endothelial growth factor (VEGF) molecule and VEGF receptors.1–3 We report the pathology and immunohistochemistry of VEGF, VEGF receptors (VEGF-R1, VEGF-R2, and VEGF-R3), and placental growth factor (PIGF) of two consecutive cases of peripapillary CNV secondary to inflammation and AMD treated by subretinal surgery.

Case Reports

Case 1

A 36-year-old man reported a central scotoma in his right eye. The patient reported bilateral posterior CNV associated with serpiginous choroiditis that had been unsuccessfully treated by systemic steroids (cyclosporin and azathioprine and intravitreal bevacizumab 1.25 mg) 3 months earlier. The left eye showed subretinal fibrosis. Best-corrected visual acuity in his right eye was 20/400 and fundus examination disclosed optic disk elevation and subretinal exudates next to the optic disk and in the macular area. Fluorescein angiography revealed a peripapillary CNV not affecting the macula. Optical coherence tomography (OCT) showed irregularly increased retinal thickness in the fovea and absence of foveal depression (Fig. 1). Because this was the patient’s only useful eye and after discussing other therapeutical options, he decided to have vitreoretinal surgery performed in his right eye.

(A) Retinography of case 1 showing peripapillary choroidal neovascularization (CNV) with subretinal exudates in the right eye and subretinal fibrosis in the left eye. (B) Fluorescein angiography (middle frame) showing leakage from the CNV. (C) Optical coherence tomography horizontal scan shows retinal thickening. (D) Retinography 2 months after CNV removal. (E) Optical coherence tomography horizontal scan after surgery shows normal retinal thickness.

Figure 1. (A) Retinography of case 1 showing peripapillary choroidal neovascularization (CNV) with subretinal exudates in the right eye and subretinal fibrosis in the left eye. (B) Fluorescein angiography (middle frame) showing leakage from the CNV. (C) Optical coherence tomography horizontal scan shows retinal thickening. (D) Retinography 2 months after CNV removal. (E) Optical coherence tomography horizontal scan after surgery shows normal retinal thickness.

Case 2

A 71-year-old man reported having a central scotoma in his left eye during the previous 2 months. Best-corrected visual acuity was 20/400 and fundus examination disclosed optic disk elevation and subretinal hemorrhage next to the optic disk and in the macular area with intraretinal exudates in his left eye. Fluorescein angiography revealed a peripapillary CNV that did not affect the macula. OCT showed retinal thickening with subretinal fluid in the peripapillary area and a central macular thickness of 254 μm (Fig. 2). The different therapeutical options were discussed with the patient, including the frequent need for intravitreal injections. In view of the surgeon’s negative personal experience with peripapillary CNV treated by intravitreal injections and the better results achieved in these cases by surgical removal of neovascular tissue, the patient decided to have surgery.

(A) Retinography of case 2 showing peripapillary choroidal neovascularization (CNV) with subretinal blood. (B) Peripapillary optical coherence tomography scan shows retinal thickening and fluid. (C) Two months after CNV removal. (D) Retinography 5 months after CNV removal. (E) Optical coherence tomography horizontal scan after surgery shows normal retinal thickness.

Figure 2. (A) Retinography of case 2 showing peripapillary choroidal neovascularization (CNV) with subretinal blood. (B) Peripapillary optical coherence tomography scan shows retinal thickening and fluid. (C) Two months after CNV removal. (D) Retinography 5 months after CNV removal. (E) Optical coherence tomography horizontal scan after surgery shows normal retinal thickness.

Both patients underwent subretinal surgery for CNV removal and the tissue was processed for immunohistochemistry. Endothelial, pigment, and glial cells were observed and new vessels were clearly identifiable in both cases (Figs. 3 and 4). New vessels were organized in clusters (Figs. 3B, 3C, and 4C), surrounded by extracellular matrix and cells. Pigment cells were observed along one of the edges of the CNV in case 1. High magnification (Fig. 3D) allowed identification of cells engulfed with pigment corresponding to retinal pigment epithelial cells. Pigment cells were observed occasionally (Fig. 4D) in case 2. A well-organized cell layer was present at the CNV edge corresponding to retinal pigment epithelial cells that had lost their pigment.

(A) Choroidal neovascularization (CNV) sections from case 1 at low magnification phase contrast microscopy (bar = 500 μm). (B to D) CNV stained with toluidine blue at higher magnification (bar = 50 μm). (B and C) Vessels and endothelial cells are visible (arrows and insets) surrounded by extracellular matrix. Notice the vascular cluster marked by *. (D) Pigment cells at one edge of the CNV that may correspond to retinal pigment epithelial cells (inset ×3).

Figure 3. (A) Choroidal neovascularization (CNV) sections from case 1 at low magnification phase contrast microscopy (bar = 500 μm). (B to D) CNV stained with toluidine blue at higher magnification (bar = 50 μm). (B and C) Vessels and endothelial cells are visible (arrows and insets) surrounded by extracellular matrix. Notice the vascular cluster marked by *. (D) Pigment cells at one edge of the CNV that may correspond to retinal pigment epithelial cells (inset ×3).

(A) Choroidal neovascularization (CNV) sections from case 2 using low magnification phase contrast microscopy (bar = 500 μm). (B to D) Photomicrographs of the same CNV stained with toluidine blue at higher magnification (bar = 50 μm). (B) CNV is formed by extracellular matrix, vessels, and cells with a well-organized monolayer of polarized cells that may correspond to transdifferentiated retinal pigment epithelial cells. (C) Vessels and endothelial cells are clearly identified (black arrows and insets) organized in clusters. (D) Pigment cells are sparse and may correspond to melanophages (inset ×3).

Figure 4. (A) Choroidal neovascularization (CNV) sections from case 2 using low magnification phase contrast microscopy (bar = 500 μm). (B to D) Photomicrographs of the same CNV stained with toluidine blue at higher magnification (bar = 50 μm). (B) CNV is formed by extracellular matrix, vessels, and cells with a well-organized monolayer of polarized cells that may correspond to transdifferentiated retinal pigment epithelial cells. (C) Vessels and endothelial cells are clearly identified (black arrows and insets) organized in clusters. (D) Pigment cells are sparse and may correspond to melanophages (inset ×3).

New vessels were clearly identified by anti-Von Willebrand factor antibody, showing well-organized and dense neovascularization in both cases (Figs. 5A and 5B).

Von Willebrand factor (VW), vascular endothelial growth factor (VEGF), and VEGF-R1 immunohistochemistry. (A and B) Von Willebrand factor immunohistochemistry combined with nuclei (DAPI) stain from cases 1 and 2, showing several vessels (arrows) and autofluorescence from pigment (stars). (C and D) VEGF-165 immunohistochemistry combined with nuclei (DAPI) stain showing absence of VEGF in case 1 (C) and clusters of vessels with VEGF in and around the vessels (gray arrows). Vessels and surrounding cells are negative for VEGF (white arrows) in other areas in case 2. (E and F) VEGF-R1 immunohistochemistry combined with nuclei (DAPI) stain showing VEGF-R1 expressed in vessels within the choroidal neovascularization (insets) in cases 1 and 2. Bar = 50 μm.

Figure 5. Von Willebrand factor (VW), vascular endothelial growth factor (VEGF), and VEGF-R1 immunohistochemistry. (A and B) Von Willebrand factor immunohistochemistry combined with nuclei (DAPI) stain from cases 1 and 2, showing several vessels (arrows) and autofluorescence from pigment (stars). (C and D) VEGF-165 immunohistochemistry combined with nuclei (DAPI) stain showing absence of VEGF in case 1 (C) and clusters of vessels with VEGF in and around the vessels (gray arrows). Vessels and surrounding cells are negative for VEGF (white arrows) in other areas in case 2. (E and F) VEGF-R1 immunohistochemistry combined with nuclei (DAPI) stain showing VEGF-R1 expressed in vessels within the choroidal neovascularization (insets) in cases 1 and 2. Bar = 50 μm.

VEGF-165 was detected only in some of the vessels from case 2 (Fig. 5C, green arrows), mostly in the cells surrounding the vascular lumen. Other cells at the edge of the CNV expressed VEGF corresponding to differentiated retinal pigment epithelial cells (Fig. 5D, star). VEGF-165 was not expressed in some areas with clusters of vessels from case 2 (Fig. 5D, white arrows).

VEGF-R1 was found in the endothelial cells of some vessels in the CNV from both patients (Figs. 5E and 5F). VEGF-R2 was not expressed in any case. VEGF-R1 was expressed in the endothelial cells in the CNV from case 2 and in the epithelial-like layer of cells bordering the edge of the CNV that could correspond to retinal pigment epithelial cells (Fig. 5F, insets). VEGF-R3 was localized in sparse organized vessels in case 1. PIGF could not be detected in our samples.

Glial fibrillary acidic protein (GFAP) positive cells were observed at the edge of the CNV and around the vessels in case 1 and at the retinal side of the CNV in case 2.

Discussion

Peripapillary CNV accounts for approximately 10% of all cases of CNV.4 This condition may resolve spontaneously or, less frequently, grow and cause submacular hemorrhages or serous macular detachment reducing visual acuity. Anti-VEGF therapies may not be effective in all cases, particularly if other pro-angiogenic factors contribute to the pathogenesis of peripapillary CNV. Surgical removal could therefore be considered as an option.

Our light microscopy findings are in agreement with previous histological reports on specimens excised from subfoveal CNV5 and peripapillary CNV1–3 that describe vascularized fibrous tissues and retinal pigment epithelial hyperplasia, frequently lining new vessels and the surface of the scars. However, probably due to the shorter evolution of our cases, the fibrotic component was not as evident as in the cases reported by other authors.

VEGF was not expressed in the CNV from case 1 that had been previously treated with bevacizumab within the previous 3 months, whereas it was found in some of the vessels and in the cells surrounding the vessels in case 2. VEGF-R1 was detected in endothelial cells in both cases and in epithelioid cells that could correspond to retinal pigment epithelial cells in the older patient. Previous analysis of macular CNV from various origins has shown that VEGF-R1 is more frequently expressed in new vessels than VEGF-R2 and that pigment cells may also express VEGF-R1.5

VEGF-R3 was found in the lesion previously treated by an anti-VEGF drug. Previous analysis of CNV in AMD has shown that VEGF-R3 can be expressed in human CNV and is up regulated after photodynamic therapy.5,6 VEGF-R3 was only expressed in the younger patient. The exact role of VEGF-R3 and its modulation by anti-VEGF treatments remain to be studied.

Otani et al. reported on immunohistological findings on CNV tissue surgically removed from patients with submacular CNV from varied etiologies.5 All CNV specimens had strong immunoreactivity for VEGF that tended to be greater in vascularized lesions and pigment-containing cells. It is likely that the absence of VEGF in case 1 was caused by the effect of intravitreal bevacizumab, even if it had occurred 3 months earlier and with no apparent clinical effect. PIGF was not detected in either case in our series, whereas in Otani et al.’s series, PIGF was present in all cases in retinal pigment epithelial and vascular cells and almost half of the pigment-containing cells and most vascular cells had strong immunoreactivity for PIGF. It has been suggested that PIGF1 and VEGF-R1 could play a role in the pathogenesis of CNV.

Regarding VEGF receptor staining, more cells were positive for VEGF-R1 than for other receptors in our series and in Otani et al.’s series. Pigment-containing cells had strong immunoreactivity for all receptor subtypes and there were more VEGF-R1–positive vessels than VEGF-R2–positive vessels in all cases in Otani et al.’s series. VEGF-R2 has been found in subfoveal CNV from patients with AMD and is known to play an important role in the proliferation of endothelial cells. However, this finding has been inconsistently reported.

VEGF-R3 was found in the lesion previously treated by an anti-VEGF drug in the younger patient. The exact role of VEGF-R3 and its modulation by anti-VEGF treatments remain to be studied. Otani et al. reported VEGF-R3 staining occurring predominantly in pigment-containing cells and less frequently in vascular cells.5

Intravitreal injections of anti-VEGF drugs in the presence of pro-angiogenic stimuli such as inflammation may trigger different pathways of neovascularization with different mediators that would be responsible for the cases of resistance to anti-angiogenic treatment. However, these facts can only be hypothesized and longer series of more homogeneous samples are required to draw such conclusions.

Our results, even if limited, show that differences may exist depending on the etiology of CNV, the patient’s age, and probably the evolution of the pathology and therapies used, suggesting that therapies directed only at controlling VEGF may be effective only in a subtype of cases and at specific evolution time points.

References

  1. Castellarin AA, Sugino IK, Nasir M, Zarbin MA. Clinicopathological correlation of an excised choroidal neovascular membrane in pseudotumour cerebri. Br J Ophthalmol. 1997;81:994–1000.
  2. Kase S, Parikh JG, Rao NA. Peripapillary subretinal neovascularization in retinoblastoma. Graefes Arch Clin Exp Ophthalmol. 2008;246:931–934.
  3. Tran HV, Bovey EH, Uffer S, Zografos L. Peripapillary choroidal neovascularization associated with melanocytoma of the optic disc: a clinicopathologic case report. Graefes Arch Clin Exp Ophthalmol. 2006;244:1367–1369.
  4. Kertes PJ. Massive peripapillary subretinal neovascularization: an indication for submacular surgery. Retina. 2004;24:219–225.
  5. Otani A, Takagi H, Oh H, et al. Vascular endothelial growth factor family and receptor expression in human choroidal neovascular membranes. Microvasc Res. 2002;64:162–169.
  6. Schmidt-Erfurth U, Schlotzer-Schrehard U, Cursiefen C, Michels S, Beckendorf A, Naumann GO. Influence of photodynamic therapy on expression of vascular endothelial growth factor (VEGF), VEGF receptor 3, and pigment epithelium-derived factor. Invest Ophthalmol Vis Sci. 2003;44:4473–4480.
Authors

From the Department of Ophthalmology (JMR-M), Albacete Medical School, Castilla La Mancha University, Albacete, Spain; the Alicante Institute of Ophthalmology (JMR-M, JAM), VISSUM, Vitreo-Retinal Unit, Alicante, Spain; the Department of Ophthalmology (JAM), Pio del Rio Hortega University Hospital, Valladolid, Spain; Centre de Recherche des Cordeliers (LJ, J-CPJ, FB-C), Université Paris Descartes, Paris, France; and the Department of Ophthalmology (FB-C), Université Paris Descartes, Paris, France.

Presented at the 11th Euretina meeting, May 29, 2011, London, United Kingdom.

Supported in part by a grant of the Spanish Ministry of Health, Instituto de Salud Carlos III, Red Temática de Investigación Cooperativa en Salud “Patología ocular del envejecimiento, calidad visual y calidad de vida” (RD07/0062).

The authors have no financial or proprietary interest in the materials presented herein.

Address correspondence to Jose Maria Ruiz-Moreno, MD, PhD, Departamento de Ciencias Médicas, Facultad de Medicina, Avenida de Almansa, 14, 02006 Albacete, Spain. E-mail: josemaria.ruiz@uclm.es

10.3928/15428877-20120308-01

Sign up to receive

Journal E-contents