June 2005
Volume 46, Issue 6
Free
Retina  |   June 2005
Anastomotic Vessels Remain Viable after Photodynamic Therapy in Primate Models of Choroidal Neovascularization
Author Affiliations
  • Mark H. Criswell
    From the Retina Service Research Laboratories, Department of Ophthalmology, Indiana University School of Medicine, Indianapolis, Indiana;
  • Thomas A. Ciulla
    From the Retina Service Research Laboratories, Department of Ophthalmology, Indiana University School of Medicine, Indianapolis, Indiana;
  • Lisa A. Lowseth
    Alcon Research, Ltd., Fort Worth, Texas; and
  • Ward Small
    Miravant Medical Technologies, Inc., Santa Barbara, California.
  • Ronald P. Danis
    From the Retina Service Research Laboratories, Department of Ophthalmology, Indiana University School of Medicine, Indianapolis, Indiana;
  • Dennis L. Carson
    Alcon Research, Ltd., Fort Worth, Texas; and
Investigative Ophthalmology & Visual Science June 2005, Vol.46, 2168-2174. doi:10.1167/iovs.04-1442
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Mark H. Criswell, Thomas A. Ciulla, Lisa A. Lowseth, Ward Small, Ronald P. Danis, Dennis L. Carson; Anastomotic Vessels Remain Viable after Photodynamic Therapy in Primate Models of Choroidal Neovascularization. Invest. Ophthalmol. Vis. Sci. 2005;46(6):2168-2174. doi: 10.1167/iovs.04-1442.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

purpose. Anastomotic vessels in exudative age-related macular degeneration (AMD) represent a serious clinical feature that reportedly does not respond well to either photocoagulation or photodynamic therapy (PDT). Anastomoses also occur in various animal models of choroidal neovascularization (CNV). In the present study, anastomotic vessels and their patency were evaluated in two primate CNV laser-trauma models after PDT, by using two novel photosensitizers.

methods. In cynomolgus (Macaca fascicularis) and squirrel (Saimiri sciureus) monkey eyes (n = 20), matrix placement of laser photocoagulation sites elicited CNV as a component of the development of fibrovascular tissue (FVT). FVT sites received PDT according to specific drug infusion and laser light treatment parameters. FVTs and anastomoses were evaluated by fundus photography, fluorescein angiography, and histologic examination.

results. Anastomoses averaged approximately 48% of FVT sites, with greatest occurrence in the macaque. Although PDT with each photosensitizer effectively produced FVT closure, both retinal vessels and anastomoses remained patent.

conclusions. Although PDT is effective in closing the choroidal neovascularization in FVT, this technique was ineffective in occluding anastomotic vessels and their associated tributaries within the mid- to proximal retina. Various factors (vascular diameter, composition, blood flow, orientation) may contribute to continued anastomotic patency. By convention, such vessels would typically be defined as chorioretinal anastomoses (CRAs); however, continuing studies suggest the possibility that these neovessels constitute dual-origin hybrids. Regardless of origin, viable anastomoses provide one potential mechanism for revascularization to occur after PDT and may help to explain why CRAs are considered a poor prognostic sign in patients with AMD.

Anastomotic vessels in the retina are commonly identified according to their presumed origin and termination points by terms that include chorioretinal (choroidal retinal) anastomoses (CRAs), which also can be labeled deep retinal vascular anomalous complexes [RVACs], 1 2 or alternatively, retinochoroidal (retinal choroidal) anastomoses (RCAs), which also can originate from retinal angiomatous proliferations (RAPs). 3 Regardless of their origin/termination or their flow direction, anastomoses occur as pathologic components of various retinopathic conditions that include age-related macular degeneration (AMD), 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 proliferative diabetic retinopathy, 17 uveitis, 18 19 20 21 22 congenital anomalies, 19 23 24 25 26 27 idiopathic juxtafoveolar retinal telangiectasis, 28 29 idiopathic central retinal vein occlusion, 30 sickle cell disease, 31 histoplasmosis, 19 toxoplasmosis, 32 angioid streaks, 33 and physical trauma 34 35 36 (for a general review, consult Kottow 37 ). CRAs in exudative age-related macular degeneration (AMD) with occult choroidal neovascularization (CNV) represent a serious clinical feature that appears to respond poorly to either photocoagulative therapy 1 10 22 38 or photodynamic therapy (PDT). 12 13  
This represents the first primate investigation to focus on anastomotic vessel viability as related to the prognosis for formation of a choroidal neovascular membrane (CNVM) and the efficacy of PDT. Anastomoses were evaluated as recurring components of CNV, elicited in two different primate CNV laser-trauma models, and these neovessels were evaluated for possible closure after PDT. 
In this report, the term “fibrovascular tissue” (FVT) is used, rather than CNV or CNVM, to describe more accurately the combined histopathologic presence of both CNV and collagenous fibroplasia, dual components that originate from the choriocapillaris at or around the laser photocoagulation sites. FVTs may remain subretinal in their development, or they can directly infiltrate the retinal layers with intermittent neovessels that continue beyond the central FVT mass and into the proximal retina. 39 These new vessels include the anastomoses examined in this investigation. Similarly, clinical detection of anastomoses can denote the physical adherence of the subretinal neovascular membrane to the neural retina and thus may contradict surgical intervention for the removal of the membrane. 28  
Methods
Anastomoses and FVTs developed in cynomolgus (Macaca fascicularis) and squirrel (Saimiri sciureus) monkey eyes after laser-induced rupture of Bruch’s membrane, as described in previous reported protocols. 39 40 41 42 43 Photocoagulation sites that exhibited FVTs subsequently received PDT. Cynomolgus monkeys were treated with the novel photosensitizer motexafin lutetium (Optrin; Pharmacyclics, Inc., Sunnyvale, CA), and squirrel monkeys were treated with the novel photosensitizer indium chloride methyl pyropheophorbide (PhotoPoint MV6401; Miravant Medical Technologies, Inc., Santa Barbara, CA). Each agent was used according to specific infusion and light treatment parameters relative to that agent, to achieve reliable CNV closure. 44 For a comparative analysis of these two primate species as CNV models, consult Criswell et al. 39  
Animals
The effects of PDT on FVT and anastomoses were evaluated in 20 eyes from young-adult cynomolgus macaque monkeys (Covance Research, Alice, TX) and from young-adult (ages, 2–4 years) squirrel monkeys (Charles River Laboratories, Houston, TX). Experiments involving the macaques were performed at the Alcon Research Laboratories (Fort Worth, TX), whereas squirrel monkey experiments were conducted at the Indiana University Laboratory Animal Resources facility (Bloomington, IN). Tissue processing and analyses for both species were performed at the Indiana University School of Medicine (Indianapolis, IN). All procedures were performed with strict adherence to the ARVO Statement for the Use of Animals and Ophthalmic and Vision Research and the guidelines for animal care and experimentation prepared by the Indiana University Institutional Animal Care and Use Committee and the Alcon Institutional Animal Care and Use Committee. 
For experimental procedures, including laser photocoagulation, ocular examination, and photography, cynomolgus monkeys (weight range, 2.3–2.5 kg) received intramuscular (IM) ketamine at 10 mg/kg, along with intravenous (IV) thiopental sodium (Pentothal; Alcon Laboratories, Fort Worth, TX) at 15 to 20 mg/kg, to supplement anesthesia. For ophthalmic procedures, topical 1.0% tropicamide and 2.5% phenylephrine hydrochloride were administered to achieve maximum pupillary dilation and cycloplegia. 
Squirrel monkeys (weight range, 600–950 g) received IM ketamine at 40 mg/kg and acepromazine at 2 mg/kg for anesthesia, along with IM atropine at 0.05 mg/kg, to minimize bronchial secretions. Maintenance amounts of this mixture (10%–15% of the original dose) were administered at 45-minute intervals, when necessary. For all ophthalmic procedures, topical 0.8% tropicamide and 2.5% phenylephrine hydrochloride were administered for pupillary dilation and cycloplegia. 
Laser Photocoagulation
After receiving anesthesia and undergoing pupillary dilation, animals were positioned before a slit lamp (Carl Zeiss Meditec, Inc., Jena, Germany) laser-delivery system. The fundus was visualized with a Goldmann-type plano fundus contact lens (Model OGFA; Ocular Instruments, Inc., Bellevue, WA) with 2.5% hydroxypropyl methylcellulose solution as a cushioning agent. 
For induction of photocoagulation sites in the cynomolgus monkeys, an argon laser (514 nm wavelength: Lambda Plus PDL II Photodynamic Laser; Coherent Inc., Santa Clara, CA) was used (390- and 455-mW power, 50-μm spot diameter, 0.1-second duration). Laser power was verified with a power meter (FieldMaster; Coherent, Inc.). In each cynomolgus eye, 16 to 20 photocoagulation sites were similarly arranged in a grid-like pattern of four rows by four to five columns within the macular region, as per criteria established by previous investigators. 40 41 42 43 45  
To elicit FVT development in squirrel monkey eyes, optimized laser parameters (650-mW power, 75-μm spot diameter, 0.05- or 0.1-second duration) were used with a solid-state diode laser (532-nm wavelength; OcuLight GL; Iris Medical Instruments, Inc., Mountain View, CA). 39 Laser power was verified with a power meter (Radiometer Model IL1400A, with Detector Head Model SPL024F; International Light, Inc., Newburyport, MA). A typical series of nine photocoagulation sites were placed within the macula, arranged in a grid-like pattern (consisting of three horizontal rows by three vertical columns). Follow-up fundus and fluorescein angiography examinations were performed on post-FVT induction day 1 and were repeated, typically, on day 28, to confirm development of FVT. 
For both species, deliberate care was taken to avoid inducing photocoagulation within the immediate vicinity of the fovea (within approximately 150 μm). Species-specific laser power settings were established and used that most reliably produced acute vapor bubbles, suggestive of Bruch’s membrane rupture. These experimental conditions also most reliably produced FVTs that emanated through the disrupted Bruch’s membrane and were capable of retinal infiltration. 
Photodynamic Therapy
At 22 days after laser induction of FVTs, cynomolgus monkeys were injected intravenously with a 0.6% solution of the hydrophilic, metallotexaphyrin-class photosensitizer motexafin lutetium (Optrin; Pharmacyclics, Inc.) at a dosage of 2 mg/kg. 46 47 At post-injection activation times within the optimal 10- to 45-minute time frame that previously had been determined for plateau accumulation of this PDT agent (Arbour JD, et al. IOVS 1999;40:ARVO Abstract 2111), 47 laser light (732 nm, 600-mW/cm2 irradiance, 1500-μm spot diameter) was delivered transcorneally (via slit lamp adapter) by a diode laser (Visulas 732s; Carl Zeiss Meditec), developed specifically for use with this photosensitizer, to designated FVT sites at a duration-dependent light dosage of 100 J/cm2. Some FVT sites did not receive PDT treatment and were designated as drug-only control sites. 
Approximately 28 days after laser induction of FVTs, squirrel monkeys were injected intravenously with the hydrophobic, pyropheophorbide-class photosensitizer indium chloride methyl pyropheophorbide (PhotoPoint; Miravant Pharmaceuticals, Inc.) at a dosage of 0.15 μmol/kg. 44 48 At postinjection activation times of +70 and +72 minutes, PDT laser light (664 nm, 500 mW/cm2 irradiance, 800 μm spot diameter) was delivered transcorneally (via slit lamp adapter) to FVT sites in the column located directly nasal to the fovea at a duration-dependent light dosage of 10 J/cm2 and to FVT sites in a column located directly temporal (middle column) to the fovea at a light dosage of 20 J/cm2. 48 FVT sites within the third, temporal-most column were designated as non-PDT control sites. All FVT sites for potential analyses and treatment were located within the species-appropriate portion of the macula where laser-trauma–induced CNV had been previously characterized. 39  
All treatment-designated central FVT sites were encompassed within the PDT beam diameter. In some instances, FVT development extended far from the central photocoagulation sites as “diffuse FVT” (consult Criswell et al. 39 ), and was not included within the PDT treatment zone. These regions of extended neovascularization were not included in the histologic tissue assessment. 
Ophthalmologic and Histologic Assessment
Cynomolgus monkeys underwent baseline and follow-up (day 22) slit lamp biomicroscopic and indirect ophthalmoscopic examinations. Animals were evaluated by fundus photography and fluorescein angiography (FA) at 23 days after laser induction of FVTs (i.e., 1 day after PDT), followed by euthanasia and histologic analysis of ocular tissues. For FA assessment, animals received 10% sodium fluorescein (0.14 mL/kg) administered intravenously (via the saphenous vein). Photocoagulation sites were viewed and photographed with a retinal camera and image-processing system (TRC-501A Retinal Camera and ImageNet 2000; Topcon Medical Systems, Inc., Paramus, NJ). 
Squirrel monkeys underwent a baseline ophthalmologic examination 7 to 14 days before induction of laser photocoagulation sites, a follow-up examination at approximately post–laser-treatment day 28 (just before PDT treatment), and a final examination 7 days after PDT (i.e., 35 days after laser induction of FVTs), followed by euthanasia and histologic analysis of ocular tissues. This assessment included fundus photography and fluorescein angiography (FA) with a fundus camera (FK-30; Carl Zeiss, Inc.) modified (with the inclusion of a +9-D, antireflective-coated spherical lens) to image the fundus in the small diameter (∼14.5 mm, axial length) eye. For early- and late-phase fluorescein angiography, 25% sodium fluorescein (0.1 mL/kg) was administered intravenously (via the saphenous vein). 
Eyes from both species were enucleated immediately after euthanasia, and eyecup preparations were fixed in 4% phosphate-buffered paraformaldehyde solution (overnight at room temperature). For each eye, a single square-shaped tissue block (∼8–10 mm/side), containing the photocoagulation sites, optic disc, and fovea, was hand sectioned from the eyecup preparation. Tissue sections were dehydrated, embedded in paraffin, serially sectioned (6-μm thickness), and stained for light microscopy by using a regressive hematoxylin procedure (to label cell nuclei), followed by eosin counterstaining (as a cytoplasm stain and to highlight macrophages). Tissue slide sections from each laser lesion site recovered were evaluated in their entirety and photographed. Histologic tissues were assessed for the presence or absence of experimentally induced FVTs and anastomoses, as well as the relative effect of PDT on these sites. Vascular occlusion was denoted by a uniform, typically pale-pink to cream-colored vessel appearance that was devoid of blood cells, whereas patent vessels contained concentrated populations of clearly identifiable erythrocytes that were bright pink-to-reddish in appearance. Quantitative comparisons of FVT and of anastomotic development between the two primate species were performed by estimating the proportional difference between two independent samples and then determining the probability of significance by t-test (assuming n = ∞). 
Results
Fundus photography and FA at approximately 3 to 5 weeks after laser photocoagulation did not reveal the clear presence of anastomoses within the fibroplasia, and indocyanine green angiography (ICGA) was not available. Nevertheless, within many FVT sites (Fig. 1) , qualitative FA observations indicated the presence of prominent retinal vessels, typically identified as arterioles, based on their origin at the disc and their onset of staining during initial fluorescein infusion. These fluorescent vessels infiltrated the central portion of the FVT site and either disappeared from view within the stained FVT mass (Figs. 1A 1E ; arrows), or they appeared to pass through the site and continued into the retina (Fig. 1C ; arrow). Hyperfluorescence, in combination with fibroplasia within the FVT mass, obstructed meaningful tracking of vessel pathways or of any collateral branching. After PDT occlusion of neovascularization within the FVT sites (Figs. 1B 1D 1F)patent and well-stained vessels were still visible within these sites, continuing into the FVT mass (Fig. 1B 1F ; arrows) and from which appeared to derive secondary collaterals which also appeared patent (Fig. 1D , arrow). Whether these vessels represented anastomoses could not be ascertained from the qualitative FA data in this study. 
From light microscopic tissue assessments, the number of FVT sites that occurred within the population of recovered laser photocoagulation sites was comparable to previously reported data. 39 By species (Table 1) , this ratio was significantly higher (t = 3.29, P < 0.0005) in the squirrel monkey (71/117 sites, 62% incidence) compared with that in the macaque (44/118 sites, 37% incidence). Histologically, anastomotic vessels were prevalent within FVT sites in both species (Fig. 2) ; however, distinguishable anastomoses were identified in 29 of the 44 (66%) FVT sites in the macaque, but in only 26 of 71 (37%) sites in the squirrel monkey. This species difference was also significant (t = 3.03, P < 0.0025). From the 36 FVT sites in each species that subsequently were treated with PDT (Table 1) , 21 (58%) macaque sites contained patent anastomoses, and 14 (39%) squirrel monkey sites revealed similar anastomoses. In this instance, however, the difference between species was not statistically significant (t = 1.61, P < 0.01). Other PDT-treated FVT sites suggested the presence of additional patent anastomotic development, but these vessels could not be reliably traced through the sequential histologic tissue sections. 
Typically, one or more vessels, emerging from the distal (choroidal) FVT membrane, continued radially or at a slightly oblique orientation through the retina (approximately 45° or greater, compared with the 90° radial plane of the retina) and merged with a retinal vessel to constitute an anastomotic formation. After PDT with either photosensitizer, closure of neovessels within the choroidal FVT membrane was reliably achieved, whereas the retinal vessels, as would be anticipated, remained open. Unexpectedly, anastomoses also appeared to be completely unaffected by PDT, and they contained discernible erythrocytes, presumably supplied by the retinal vasculature. 
In this study, some of the anastomoses seemed to derive from the distally infiltrating FVT mass (Figs. 2H 2I ; small arrows), and they appeared to continue proximally to the retinal vasculature; however, this is only a qualitative observation. There were also variations in anastomotic vessel diameter that suggest possible differences in source and flow direction. Vessels that appeared to be CRAs seemed to originate from the FVT membrane and possessed smaller diameters (approximately 5 μm), characteristic of the largest-diameter vessels within the principal FVT mass. 
In many other instances (Figs. 2A 2B 2C 2D 2E 2F 2G) , proximal retinal vessels within the laser photocoagulation/PDT treatment zone were noticeably larger in their vascular diameters (Figs. 2D 2E 2F ; large arrows and asterisks), measuring approximately two to five times (10–75 μm) the diameter size of normal retinal vessels (5–15 μm) that lay outside this region. Often (Figs. 2A 2B 2C 2F ; large arrows and small asterisks), from one to three primary anastomotic collaterals (“trunks”) could be seen branching from these enlarged retinal vessels within the ganglion cell layer (up to several hundred micrometers away from the photocoagulation/FVT site), and they continued distally at acute angles along and through the retinal layers to reach the FVT membrane. Two or three primary trunk vessels often approached the FVT site from different (opposite) directions (large white arrows compared with small asterisks; Figs. 2A 2B 2C ) and sometimes merged into a common junction within the FVT membrane. These primary anastomotic trunks also exhibited atypically large diameters of approximately 10 to 18 μm. Smaller, but still viable, vessels approximately 5 μm in diameter (Figs. 2D 2G ; small arrows) occasionally derived from these larger vessels and from within the FVT membrane, branching tangentially at distal and proximal angles from the primary vessel and providing additional infiltration to the FVT and surrounding retinal layers. Pigmented macrophages were often observed just adjacent to these vessels (e.g., Figs. 2G 2H ; arrowheads). Up to the principal FVT membrane, neither primary nor secondary anastomotic neovessels appeared to be occluded by PDT. 
Discussion
Both primate species demonstrated the frequent occurrence of anastomoses at FVT sites with a mean incidence of 48%. By comparison, the percentage of laser trauma sites that elicited FVT development was significantly greater in the squirrel monkey than in the macaque (62% vs. 37%); conversely, the percentage of FVT sites demonstrating anastomoses in the macaque was significantly greater than in the squirrel monkey (66% vs. 37%). 
Although PDT may effectively close neovascularization in the choroid and expanding CNV membrane, this study provides evidence that anastomoses and their branches in the middle and proximal retina can remain open after PDT. Possible reasons for continued anastomotic patency may include (1) dissemination and mediation of vessel orientation (radial versus tangential) relative to the incident angle of the PDT laser beam; (2) innate differences between retinal and choroidal vessels; (3) relatively higher retinal blood flow rate that, in the proximal retina, could locally decrease photosensitizer concentration; (4) sustained arterial pressure of retinal blood vessels that could result in distension of both retinal and anastomotic trunks after PDT occlusion distally of smaller neovessels in the FVT membrane and choriocapillaris; and (5) PDT, although effective in occluding smaller diameter neovessels of choroidal origin (up to approximately 5 μm diameter), may not be effective in blocking larger diameter blood vessels (5 μm diameter and larger). Viable anastomotic vessels could contribute to the subsequent revascularization of the neovascular membrane region after PDT. Additional studies to evaluate later time points after PDT eventually will determine whether anastomoses demonstrate long-term patency and continued angiogenesis. 
CRAs are considered a poor prognostic sign for laser treatment of CNV in general. 10 12 Moreover, Kuhn et al. 7 reported that ICG-guided photocoagulation of anastomoses in patients with AMD did not achieve effective closure. Mantel and Zografos 14 have indicated that successful CRA closure by selective photocoagulation may be beneficial in the overall treatment of AMD, whereas Silva et al. 16 have concluded that selective treatment of CRAs by PDT can result in stabilization or even improvement in visual acuity in most patients at 1 year after treatment. 
Until recently, anastomoses have been generally considered a relatively infrequent sign of late-stage, exudative AMD, owing in part to the fact that they are difficult to visualize, particularly with FA. Wolf and Goldberg 49 indicated in a previous primate investigation that CRAs did not become apparent on FA until at least 7 weeks after laser photocoagulation. Why this should happen is unclear, but it may be related to spontaneous regression of neovascularization that commences in the primate retina at approximately 5 to 8 weeks after lesion induction, thereby permitting better identification of larger anastomotic vessels. Collagenous fibroplasia surrounding the anastomotic vessels also may partially mask their presence. Moreover, anastomoses may appear as focal sites (“hot spots”) 3 of hyperfluorescence due to their characteristically high incidence angle of orientation through the retina, making them difficult to identify. ICGA, in conjunction with scanning laser ophthalmoscopy, can improve the analysis and localization of anastomotic “hot spots” and may provide certain technical advantages for visualizing these vessels. 3 7 8 10 11 12 13 50 Optical coherence tomography (OCT) may best reveal the clinical presence of anastomoses in radial (cross-sectional) views of the retina. 13  
With these improved examination techniques, the clinical presence of anastomoses is now thought to be more prevalent in patients with exudative AMD than has been previously reported. CRAs have been shown to precede or to occur in tandem with more generalized neovascularization. 12 29 The occurrence of CRAs is now regarded as a common feature, to even a primary manifestation, of classic CNV in exudative AMD. 10 13 Kuhn et al. 7 reported that one or more CRAs were visualized (on FA or ICGA) in 27% (n = 186 eyes) of 182 patients with early exudative AMD; Slakter et al. 10 documented that CRAs occurred in 21% of exudative AMD eyes (n = 150), with 71% of these eyes having exhibited multiple vessels; and Axer-Siegel et al. 11 indicated that anastomoses were visualized (on FA and ICGA) in conjunction with CNV in 28% of eyes (n = 205) in 153 AMD patients with occult CNV. In patients with unilateral AMD, the initial appearance of CRAs in the non-AMD-diagnosed eye now merits increased attention as a potential harbinger for contralateral neovascularization in this eye (Haddad WM, et al. IOVS 2003;44:ARVO E-Abstract 4991). 
Although histologic findings in this investigation suggest the possible presence of multiple anastomoses at FVT sites, the presence of multiple anastomoses also has been discovered at FVT sites in the rat laser trauma model (Criswell et al., unpublished data, 2005). In rat, confocal microscopic analyses of FITC-albumin labeled FVT sites have revealed the presence of multiple anastomoses (commonly eight or more per site), even though only one or two anastomoses may be evident during light microscopic examination of radial tissue sections, and no anastomoses were observed during fundus or FA examinations. 
Qualitative evidence in this study further suggests that anastomoses may simultaneously originate from both choroidal and retinal vascular sources and unify in and around the central FVT membrane. Some anastomoses probably represent CRAs, given that laser photocoagulation initiated neovascularization and development of FVT that originated from the choriocapillaris. However, several lines of evidence indicate that such an assumption may not apply uniformly to all identified anastomoses. In previous work, investigators have demonstrated that it is not uncommon for more than one vascular direction (choroidal to retinal versus retinal to choroidal) as well as type (artery to vein, artery to artery, vein to vein) of anastomotic permutation to evolve. 11 19 32 51 52 Moreover, the larger-diameter anastomoses observed proximally may represent RCAs of retinal origin, rather than CRAs of choroidal origin. The enlarged diameters of the retinal feeder vessels and of certain proximal anastomoses suggest the possibility of a retinal arterial origin for these particular anastomoses, with colligation of these vessels and some of their smaller collateral tributaries within the FVT membrane. Human clinical evidence also suggests the possibility of dual neovascular development. 4 53 Alternatively, Yannuzzi et al. 3 have proposed that neovascularization can, in a subset of patients with AMD, originate as a primary manifestation of the retina vasculature (RAP) and later may evolve into an RCA with choroidal vessels. 
The angiogenic potential for choroidal and retinal vascular networks to respond mutually in forming anastomoses also exists. Recent and continuing research (Hu WZ, et al. IOVS 2004;45:ARVO E-Abstract 1844) has provided evidence that laser trauma to the retinal pigment epithelium-choriocapillaris not only upregulates active growth factor expression to evoke FVT development in the choriocapillaris, but also differentially increases expression of certain precursor and active growth factors in the proximal retina, thereby increasing the possible initiation of angiogenic events within the retinal vascular layer. Clinical evidence supports this finding. Wallow et al. 38 reported that in a patient with AMD, a new neovascular membrane had developed, which was supplied by the retinal vasculature, approximately 5 months after photocoagulative treatment of the original CNV membranes. Similarly, patients with early exudative AMD who were treated with laser photocoagulative therapy for vascularized retinal pigment epithelium detachments showed revascularization of the sites by anastomotic vessels of retinal origin (RCAs). 7 10 14  
In conclusion, this is the first primate study to examine continuing anastomotic viability in conjunction with CNV and PDT efficacy. CNV in diseases such as exudative AMD can commonly include anastomotic vessel development as an additional consequence of angiogenesis. Typical photodynamic therapy, while designed to achieve choroidal neovascular closure, may not be adequate to occlude anastomotic vessels. The continued patency of anastomoses provides one possible neovascular source for revascularization and may help to explain why the visualization of anastomoses is considered such a poor prognostic sign. Careful identification and selective highly localized treatment of anastomotic vessels by PDT or photocoagulation, in addition to standard PDT for treatment of CNVMs, may be beneficial to a patient’s long-term visual outcome. The combined use of PDT and antiangiogenic therapies may also contribute to a more favorable long-term management strategy (Haddad WM, et al. IOVS 2004;45:ARVO E-Abstract 3163). 
 
Figure 1.
 
Fluorescein angiography of 3 individual photocoagulation sites (A, C, E) that elicited development of FVT, and those same sites again (B, D, F) after PDT, which yielded partial (B) to essentially total (F) occlusion of neovascularization. Nevertheless, each example appears to suggest qualitatively the possibility of a still viable anastomosis within each FVT site. In (A), a large retinal vessel and numerous smaller collaterals were discernible within the central photocoagulation site (arrow); whereas in (B), PDT occluded the neovessels (arrow), but the larger vessel still appeared patent and its continuation into the central FVT mass actually became more evident. In (C), a large retinal vessel bifurcated within the central region of the site; however after PDT (D), one branch (arrow) appeared to become more prominent in its staining than that observed (C) before treatment, and tertiary vessels appeared to emanate from diffuse staining within this region. In (E), two retinal vessels appeared to form a junction within the central portion of the site, from which a single branch emerged that appeared to continue approximately 25 μm into the central FVT mass. After PDT (F) this same vessel (still patent) appeared to continue further (∼50 μm) into the central portion of the treated FVT site. Scale bar, 75 μm.
Figure 1.
 
Fluorescein angiography of 3 individual photocoagulation sites (A, C, E) that elicited development of FVT, and those same sites again (B, D, F) after PDT, which yielded partial (B) to essentially total (F) occlusion of neovascularization. Nevertheless, each example appears to suggest qualitatively the possibility of a still viable anastomosis within each FVT site. In (A), a large retinal vessel and numerous smaller collaterals were discernible within the central photocoagulation site (arrow); whereas in (B), PDT occluded the neovessels (arrow), but the larger vessel still appeared patent and its continuation into the central FVT mass actually became more evident. In (C), a large retinal vessel bifurcated within the central region of the site; however after PDT (D), one branch (arrow) appeared to become more prominent in its staining than that observed (C) before treatment, and tertiary vessels appeared to emanate from diffuse staining within this region. In (E), two retinal vessels appeared to form a junction within the central portion of the site, from which a single branch emerged that appeared to continue approximately 25 μm into the central FVT mass. After PDT (F) this same vessel (still patent) appeared to continue further (∼50 μm) into the central portion of the treated FVT site. Scale bar, 75 μm.
Table 1.
 
Incidence of Patent Anastomotic Vessel Development within FVT Sites
Table 1.
 
Incidence of Patent Anastomotic Vessel Development within FVT Sites
Monkey Species Cynomolgus Macaque (M. fascicularis) Squirrel (S. sciureus) Totals
Total eyes evaluated 8 12 20
Eyes receiving PDT treatment 8 6 14
Total laser lesion sites induced 120 117 237
Total lesion sites recovered 118 117 235
Total FVT sites (percentage of total lesion sites recovered) 44 (37) 71 (62) 115 (49)
Total anastomoses (percentage of total FVT sites) 29 (66) 26 (37) 55 (48)
PDT-treated infiltrative FVT sites 36 36 72
Patent anastomoses in PDT-FVT sites (percentage of PDT-treated FVT sites) 21 (58) 14 (39) 35 (49)
Additional unconfirmed anastomoses (percentage of PDT-treated infiltrative FVT sites) 4 (11) 6 (17) 10 (14)
Figure 2.
 
Serial (AE) and single (FI) radial views of chorioretinal anastomoses (CRAs) at six FVT sites in histologic tissues from the macaque and squirrel monkeys. At these and other sites assessed, PDT effectively closed vessels within the choriocapillaris and distal FVT membrane, whereas patency of anastomotic vessels located in the proximal portion of the retina (AI, large arrows; D, small arrows; AC, F, small asterisks) continued (as defined respectively by the absence or presence of erythrocytes within these vessels). Retinal blood vessels in the vicinity of the photocoagulation-FVT site (D, E, large asterisks) were enlarged (up to five times their normal diameters); these vessels continued to supply blood to viable CRAs by means of large trunk vessels (AG, large arrows; AC, F, small asterisks) and by their disseminating branch vessels (D, small arrows). Macrophages were commonly observed in close conjunction with newly developed vessels within the retina (G, H, arrowheads). Scale bar, 100 μm.
Figure 2.
 
Serial (AE) and single (FI) radial views of chorioretinal anastomoses (CRAs) at six FVT sites in histologic tissues from the macaque and squirrel monkeys. At these and other sites assessed, PDT effectively closed vessels within the choriocapillaris and distal FVT membrane, whereas patency of anastomotic vessels located in the proximal portion of the retina (AI, large arrows; D, small arrows; AC, F, small asterisks) continued (as defined respectively by the absence or presence of erythrocytes within these vessels). Retinal blood vessels in the vicinity of the photocoagulation-FVT site (D, E, large asterisks) were enlarged (up to five times their normal diameters); these vessels continued to supply blood to viable CRAs by means of large trunk vessels (AG, large arrows; AC, F, small asterisks) and by their disseminating branch vessels (D, small arrows). Macrophages were commonly observed in close conjunction with newly developed vessels within the retina (G, H, arrowheads). Scale bar, 100 μm.
The authors thank Russell L. Schmidt, Director, Indiana University Laboratory Animal Resources (Bloomington, Indiana), for cooperation and assistance in the project; Tiffany E. Hill, Lisa L. Bird-Turner, and Robert S. Flack for technical expertise and assistance; and Richard A. Miller, MD, (Pharmacyclics Inc., Sunnyvale, CA) for his cooperation in the preparation of the manuscript. 
HartnettME, WeiterJJ, StaurenghiG, ElsnerAE. Deep retinal vascular anomalous complexes in advanced age-related macular degeneration. Ophthalmology. 1996;103:2042–2053. [CrossRef] [PubMed]
LafautBA, AisenbreyS, VandenBroeche C, Bartz-SchmidtKU. Clinicopathological correlation of deep retinal vascular anomalous complex in age related macular degeneration. Br J Ophthalmol. 2000;84:1269–1274. [CrossRef] [PubMed]
YannuzziLA, NegrãoS, IidaT, et al. Retinal angiomatous proliferation in age-related macular degeneration. Retina. 2001;21:416–434. [CrossRef] [PubMed]
GreenWR, GassDM. Senile disciform degeneration of the macula. Arch Ophthalmol. 1971;86:487–494. [CrossRef] [PubMed]
CialdiniAP, JalkhAE, TrempeCL, WeiterJJ, NasrallahFP, RoodhooftJM. Photocoagulation of chorioretinal anastomoses in far-advanced age-related macular degeneration. Ophthalmic Surg. 1989;20:316–320. [PubMed]
FryczkowskiAW, ShermanMD, AllinsonRM, PayneCM. Chorioretinal anastomoses in age-related macular degeneration. Ann Ophthalmol. 1989;21:370–374,378. [PubMed]
KuhnD, MeunierI, SoubraneG, CoscasG. Imaging of chorioretinal anastomoses in vascularized retinal pigment epithelium detachments. Arch Ophthalmol. 1995;113:1392–1398. [CrossRef] [PubMed]
SchneiderU, GeliskenF, KreissigI. Retinal choroidal anastomosis in classic choroidal neovascularization demonstrated by indocyanine green angiography. Acta Ophthalmol Scand. 1995;73:450–452. [PubMed]
DingPC, ChenMT, TsengSC. Choroidal retinal anastomosis in age-related macular degeneration as demonstrated by fluorescein angiography: a case report. Kaohsiung J Med Sci. 2000;16:266–268. [PubMed]
SlakterJS, YannuzziLA, SchneiderU, et al. Retinal choroidal anastomoses and occult choroidal neovascularization in age-related macular degeneration. Ophthalmology. 2000;107:742–753. [CrossRef] [PubMed]
Axer-SiegelR, BourlaD, PrielE, YassurY, WeinbergerD. Angiographic and flow patterns of retinal choroidal anastomoses in age-related macular degeneration with occult choroidal neovascularization. Ophthalmology. 2002;109:1726–1736. [CrossRef] [PubMed]
GassJD, AgarwalA, LavinaAM, TawansyKA. Focal inner retinal hemorrhages in patients with drusen: an early sign of occult choroidal neovascularization and chorioretinal anastomosis. Retina. 2003;23:741–751. [CrossRef] [PubMed]
KusserowC, MichelsS, Schmidt-ErfurthU. Chorioretinal anastomosis as unfavourable prognostic factor during photodynamic therapy (in German). Ophthalmologe. 2003;100:197–202. [CrossRef] [PubMed]
MantelI, ZografosL. Photocoagulation of the retinal feeder vessels of a chorioretinal anastomosis in age-related degeneration (in French). J Fr Ophthalmol. 2003;26:493–497.
SpaideRF. Fundus autofluorescence and age-related macular degeneration. Ophthalmology. 2003;110:392–399. [CrossRef] [PubMed]
SilvaRM, Faria De AbreuJR, TravassosA, Cunha-VazJG. Stabilization of visual acuity with photodynamic therapy in eyes with chorioretinal anastomoses. Graefes Arch Clin Exp Ophthalmol. 2004;242:368–376. [CrossRef] [PubMed]
GalinosSO, McMeelJW, TrempeCL, SchepensCL. Chorioretinal anastomoses after argon laser photocoagulation. Am J Ophthalmol. 1976;82:241–245. [CrossRef] [PubMed]
KennedyJE, WiseGN. Retinochoroidal vascular anastomosis in uveitis. Am J Ophthalmol. 1971;71:1221–1225. [CrossRef] [PubMed]
SlusherMM, TylerME. Choroidoretinal vascular anastomoses. Am J Ophthalmol. 1980;90:217–222. [CrossRef] [PubMed]
PotterJW. Vascular anastomoses in ocular toxoplasmosis. J Am Optom Assoc. 1982;53:549–552. [PubMed]
OberRR, SmithRE, RyanSJ. Subretinal neovascularization in the Vogt-Koyanagi-Harada syndrome. Int Ophthalmol. 1983;6:225–234. [CrossRef] [PubMed]
WallowI, JohnsK, BarryP, ChandraS, BindleyC. Chorioretinal and choriovitreal neovascularization after photocoagulation for proliferative diabetic retinopathy: a clinicopathologic correlation. Ophthalmology. 1985;92:523–532. [CrossRef] [PubMed]
AwanKJ. Familial chorioretinal vascular anastomoses and congenital cataracts. J Pediatr Ophthalmol Strabismus. 1980;17:384–388. [PubMed]
van NouhuysCE, DeutmanAF. Congential bilateral arterial anastomosis between the choroid and peripheral retina. Am J Ophthalmol. 1980;90:154–159. [CrossRef] [PubMed]
CohenSY, QuentelG, CuiberteauB, CoscasGJ. Retinal vascular changes in congenital hypertrophy of the retinal pigment epithelium. Ophthalmology. 1993;100:471–474. [CrossRef] [PubMed]
GaddieIB, BloomSM. Idiopathic retinal venous-choroidal anastomosis. Retina. 1998;18:483–485. [CrossRef] [PubMed]
LeysA, GilbertHD, Van De SompelW, et al. Familial spastic paraplegia and maculopathy with juxtafoveolar retinal telangiectasis and subretinal neovascularization. Retina. 2000;20:184–189. [CrossRef] [PubMed]
BergerAS, McCuenBW, II, BrownGC, BrownlowRL, Jr. Surgical removal of subfoveal neovascularization in idiopathic juxtafoveolar retinal telangiectasis. Retina. 1997;17:94–98. [CrossRef] [PubMed]
EngelbrechtNE, AabergTM, Jr, SungJ, LewisML. Neovascular membranes associated with idiopathic juxtafoveolar telangiectasis. Arch Ophthalmol. 2002;120:320–324. [CrossRef] [PubMed]
TakahashiK, MuraokaK, KishiS, ShimizuK. Formation of retinochoroidal collaterals in central retinal vein occlusion. Am J Ophthalmol. 1998;126:91–99. [CrossRef] [PubMed]
NagpalKC, GoldbergMF, RabbMF. Ocular manifestations of sickle hemoglobinopathies. Surv Ophthalmol. 1977;21:391–411. [CrossRef] [PubMed]
OwensPL, GoldbergMF, BusseBJ. Prospective observation of vascular anastomoses between the retina and choroid in recurrent toxoplasmosis. Am J Ophthalmol. 1979;88:402–405. [CrossRef] [PubMed]
SecrétanM, ZografosL, GuggisbergD, PiquetB. Chorioretinal vascular abnormalities associated with angioid streaks and pseudoxanthoma elasticum. Arch Ophthalmol. 1998;116:1333–1336. [CrossRef] [PubMed]
GoldbergMF. Chorioretinal vascular anastomoses after blunt trauma to the eye. Am J Ophthalmol. 1976;82:892–895. [CrossRef] [PubMed]
GoldbergMF. Chorioretinal vascular anastomoses after perforating trauma to the eye. Am J Ophthalmol. 1978;85:171–173. [CrossRef] [PubMed]
FraserEA, BarrDB. Chorioretinal arterial and venous anastomoses as a result of blunt trauma. Eye. 2004;18:336–338. [CrossRef] [PubMed]
KottowM. Vascular anastomoses at the posterior pole of the eye (in German). Klin Monatsbl Augenheilkd. 1979;175:513–520. [PubMed]
WallowIH, MyersFL, KimYM, BindleyC. Subretinal new vessels after krypton laser photocoagulation. Arch Ophthalmol. 1985;103:1844–1848. [CrossRef] [PubMed]
CriswellMH, CiullaTA, HillTE, et al. The squirrel monkey: characterization of a New-World primate model of experimental choroidal neovascularization and comparison with the macaque. Invest Ophthalmol Vis Sci. 2004;45:625–634. [CrossRef] [PubMed]
MillerJW, WalshAW, KramerM, et al. Photodynamic therapy of experimental choroidal neovascularization using lipoprotein-delivered benzoporphyrin. Arch Ophthalmol. 1995;113:810–818. [CrossRef] [PubMed]
HusainD, KramerM, KennyAG, et al. Effects of photodynamic therapy using verteporfin on experimental choroidal neovascularization and normal retina and choroid up to 7 weeks after treatment. Invest Ophthalmol Vis Sci. 1999;40:2322–2331. [PubMed]
RyanSJ. The development of an experimental model of subretinal neovascularization in disciform macular degeneration. Trans Am Ophthalmol Soc. 1979;77:707–745. [PubMed]
RyanSJ. Subretinal neovascularization, natural history of an experimental model. Arch Ophthalmol. 1982;100:1804–1809. [CrossRef] [PubMed]
CiullaTA, CriswellMH, DanisRP, SnyderWJ, SmallW. Evaluation of PhotoPoint photosensitizer MV6401, indium chloride methyl pyropheophorbide as a photodynamic therapy agent in primate choriocapillaris and laser-induced choroidal neovascularization. Retina. 2004;24:521–529. [CrossRef] [PubMed]
TolentinoMJ, HusainD, TheodosiadisP, et al. Angiography of fluoresceinated anti-vascular endothelial growth factor and dextrans in experimental choroidal neovascularization. Arch Ophthalmol. 2000;118:78–84. [CrossRef] [PubMed]
BlumenkranzMS, WoodburnKW, QingF, VerdoonerS, KesselD, MillerR. Lutetium texaphyrin (Lu-Tex): a potential new agent for ocular fundus angiography and photodynamic therapy. Am J Ophthalmol. 2000;129:353–362. [CrossRef] [PubMed]
WoodburnKW, EngelmanCJ, BlumenkranzMS. Photodynamic therapy for choroidal neovascularization. Retina. 2002;22:391–405. [CrossRef] [PubMed]
CiullaTA, CriswellMH, SnyderWJ, SmallW, IV. Photodynamic therapy with PhotoPoint photosensitizer MV6401, indium chloride methyl pyropheophorbide, achieves selective closure of rat corneal neovascularization and rabbit choriocapillaris. Br J Ophthalmol. 2005;89:113–119. [CrossRef] [PubMed]
WolfED, GoldbergMF. Chorioretinal vascular anastomoses resulting from photocoagulation in cynomolgus monkeys. Ophthalmic Surg. 1980;11:30–38. [PubMed]
BrowningDJ. Fundus photographic, fluorescein angiographic, and indocyanine green angiographic signs in successful laser chorioretinal venous anastomosis for central retinal vein occlusion. Ophthalmology. 1999;106:2261–2268. [CrossRef] [PubMed]
PolettiJ. Acquired chorio-retinal anastomoses (in French). J Fr Ophtalmol. 1980;3:737–741. [PubMed]
DoftBH. Choroidoretinal vascular anastomosis. Arch Ophthalmol. 1983;101:1053–1054. [CrossRef] [PubMed]
ChaumE, GreenwaldMA. Retinochoroidal anastomoses and a choroidal neovascular membrane in a macular exudate following treatment for retinal macroaneurysms. Retina. 2002;22:363–366. [CrossRef] [PubMed]
Figure 1.
 
Fluorescein angiography of 3 individual photocoagulation sites (A, C, E) that elicited development of FVT, and those same sites again (B, D, F) after PDT, which yielded partial (B) to essentially total (F) occlusion of neovascularization. Nevertheless, each example appears to suggest qualitatively the possibility of a still viable anastomosis within each FVT site. In (A), a large retinal vessel and numerous smaller collaterals were discernible within the central photocoagulation site (arrow); whereas in (B), PDT occluded the neovessels (arrow), but the larger vessel still appeared patent and its continuation into the central FVT mass actually became more evident. In (C), a large retinal vessel bifurcated within the central region of the site; however after PDT (D), one branch (arrow) appeared to become more prominent in its staining than that observed (C) before treatment, and tertiary vessels appeared to emanate from diffuse staining within this region. In (E), two retinal vessels appeared to form a junction within the central portion of the site, from which a single branch emerged that appeared to continue approximately 25 μm into the central FVT mass. After PDT (F) this same vessel (still patent) appeared to continue further (∼50 μm) into the central portion of the treated FVT site. Scale bar, 75 μm.
Figure 1.
 
Fluorescein angiography of 3 individual photocoagulation sites (A, C, E) that elicited development of FVT, and those same sites again (B, D, F) after PDT, which yielded partial (B) to essentially total (F) occlusion of neovascularization. Nevertheless, each example appears to suggest qualitatively the possibility of a still viable anastomosis within each FVT site. In (A), a large retinal vessel and numerous smaller collaterals were discernible within the central photocoagulation site (arrow); whereas in (B), PDT occluded the neovessels (arrow), but the larger vessel still appeared patent and its continuation into the central FVT mass actually became more evident. In (C), a large retinal vessel bifurcated within the central region of the site; however after PDT (D), one branch (arrow) appeared to become more prominent in its staining than that observed (C) before treatment, and tertiary vessels appeared to emanate from diffuse staining within this region. In (E), two retinal vessels appeared to form a junction within the central portion of the site, from which a single branch emerged that appeared to continue approximately 25 μm into the central FVT mass. After PDT (F) this same vessel (still patent) appeared to continue further (∼50 μm) into the central portion of the treated FVT site. Scale bar, 75 μm.
Figure 2.
 
Serial (AE) and single (FI) radial views of chorioretinal anastomoses (CRAs) at six FVT sites in histologic tissues from the macaque and squirrel monkeys. At these and other sites assessed, PDT effectively closed vessels within the choriocapillaris and distal FVT membrane, whereas patency of anastomotic vessels located in the proximal portion of the retina (AI, large arrows; D, small arrows; AC, F, small asterisks) continued (as defined respectively by the absence or presence of erythrocytes within these vessels). Retinal blood vessels in the vicinity of the photocoagulation-FVT site (D, E, large asterisks) were enlarged (up to five times their normal diameters); these vessels continued to supply blood to viable CRAs by means of large trunk vessels (AG, large arrows; AC, F, small asterisks) and by their disseminating branch vessels (D, small arrows). Macrophages were commonly observed in close conjunction with newly developed vessels within the retina (G, H, arrowheads). Scale bar, 100 μm.
Figure 2.
 
Serial (AE) and single (FI) radial views of chorioretinal anastomoses (CRAs) at six FVT sites in histologic tissues from the macaque and squirrel monkeys. At these and other sites assessed, PDT effectively closed vessels within the choriocapillaris and distal FVT membrane, whereas patency of anastomotic vessels located in the proximal portion of the retina (AI, large arrows; D, small arrows; AC, F, small asterisks) continued (as defined respectively by the absence or presence of erythrocytes within these vessels). Retinal blood vessels in the vicinity of the photocoagulation-FVT site (D, E, large asterisks) were enlarged (up to five times their normal diameters); these vessels continued to supply blood to viable CRAs by means of large trunk vessels (AG, large arrows; AC, F, small asterisks) and by their disseminating branch vessels (D, small arrows). Macrophages were commonly observed in close conjunction with newly developed vessels within the retina (G, H, arrowheads). Scale bar, 100 μm.
Table 1.
 
Incidence of Patent Anastomotic Vessel Development within FVT Sites
Table 1.
 
Incidence of Patent Anastomotic Vessel Development within FVT Sites
Monkey Species Cynomolgus Macaque (M. fascicularis) Squirrel (S. sciureus) Totals
Total eyes evaluated 8 12 20
Eyes receiving PDT treatment 8 6 14
Total laser lesion sites induced 120 117 237
Total lesion sites recovered 118 117 235
Total FVT sites (percentage of total lesion sites recovered) 44 (37) 71 (62) 115 (49)
Total anastomoses (percentage of total FVT sites) 29 (66) 26 (37) 55 (48)
PDT-treated infiltrative FVT sites 36 36 72
Patent anastomoses in PDT-FVT sites (percentage of PDT-treated FVT sites) 21 (58) 14 (39) 35 (49)
Additional unconfirmed anastomoses (percentage of PDT-treated infiltrative FVT sites) 4 (11) 6 (17) 10 (14)
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×