Polypoidal Choroidal Vasculopathy

Abstract

Purpose of Review

The goal of this paper is to review the recent literature of polypoidal choroidal vasculopathy (PCV) and provide an update on the epidemiology, pathophysiology, clinical findings, and management.

Recent Findings

Although indocyanine-green angiography (ICGA) is still the gold standard for diagnosis of PCV, the use of en face optical coherence tomography (OCT) and OCT angiography are useful tools in the diagnosis of PCV. Studies demonstrate superior treatment outcomes with combination photodynamic therapy (PDT) and anti-vascular endothelial growth factor (VEGF) therapy.

Summary

PCV is a disease most commonly in Asians and African-Americans and presents with an orange-red nodule in the macula or the peripapillary region. While ICGA remains the most accurate method to diagnose PCV, newer non-invasive imaging modalities (eg. OCT-A and en face OCT) can be used to identify PCV lesions. The combination of PDT and anti-VEGF therapy is superior to either monotherapy. Future studies of OCT modalities and other anti-VEGF agents will be important in guiding PCV diagnosis and management, respectively.

INTRODUCTION

Polypoidal choroidal vasculopathy (PCV), a retinal disorder with choroid vascular abnormalities, was first presented at the Macula Society Meeting in 1982 by Yannuzzi et. al.(1) as a peculiar idiopathic hemorrhagic disorder involving the macula and characterized by recurrent subretinal pigment epithelial bleeding. Since then, advances in PCV research has expanded our understanding of the disease processes, and allowed for better detection and earlier diagnosis. In 2005, the Japanese Study Group developed guidelines for the diagnosis of PCV based on fundus examination and indocyanine green (ICG) angiography: 1)protruding elevated orange-red lesions observed on fundus examination and 2) characteristic polypoidal lesions seen on ICGA.(2) Despite these advances in diagnosis of PCV, estimates from hospital or clinic based cross-sectional studies found a prevalence of 22–62% among Asians(3–11) and 8–13% in Caucasians, who were initially diagnosed with neovascular age-related macular degeneration (NV-AMD).(12) The purpose of this article is to discuss and provide an updated review of the understanding of PCV, including the epidemiology, clinical and imaging findings, pathophysiology, and current management strategies for PCV.

EPIDEMIOLOGY

PCV has been reported in a patients as young as 21 years, but typically occurs in older patients, with a mean presentation age of 68 years.(13–15) The Beijing Eye Study estimated the PCV prevalence in Chinese patients to be 0.3% using a combination of clinical exam and optical coherence tomography (OCT). PCV is more prevalent in Asian and black patients.(12,16) Among Asians, PCV more commonly occurs in males (63–88%), affects the macular region (62–94%), and is unilateral (72–%).(15) In Caucasians, PCV is more common in women (52–65%) and bilateral (21–55%).(15) Systemic and ocular risk factors associated with PCV include systemic hypertension, elevated C-reactive protein, hyperhomocysteinuria, cigarette smoking, central serous chorioretinopathy, focal choroidal excavation, and reticular pseudodrusen.(17–27)

CLINICAL FINDINGS

On fundus examination, PCV classically presents with an orange-red nodule in the macula or peripapillary region (Figure 1A) and is associated with serosanguinous pigment epithelial detachments (PEDs) without associated drusen.(1) However, drusen may be seen in Caucasian patients with PCV.(15) Polypoidal lesions are often at the notch in the borders of large PED.(28) RPE tears and/or RPE microrips may be seen at the margins of PEDs. PCV can also present with multiple, recurrent subretinal hemorrhages and fluid associated with an abnormal branching vessel network (BVN) with terminal aneurysmal dilations, resembling a “polyp-like” structure. Chang et al. found 49% of patients with presumed diagnosis of NV-AMD diagnosed as having PCV.(4) Bilateral involvement was significantly lower in PCV (8.2%) compared to NV-AMD (33.3%). Of 3468 subjects in the Beijing Eye Study, 18 eyes were found to have PCV and only 1 of 17 patients (6%) in the study showed bilateral involvement.(10) This finding contrasted with previous hospital-based studies from 1990s and 2000s that showed 14% of Japanese patients and 32% of European patients developed bilateral disease.(15) Chronic cases of PCV often have intraretinal lipid exudates surrounding the vascular lesions.(15,28,29) After resolution of the serosanguinous fluid, patients may develop RPE hyperplasia, subretinal fibrosis, and/or atrophic degeneration.(9,15)

Figure 1.

A) Color fundus photograph of a patient with treatment-naïve polypoidal choroidal vasculopathy (PCV) in the left eye. B) Fluorescein angiography demonstrating classic choroidal neovascularization (CNV). C) Indocyanine green angiography (ICGA) demonstrating classic CNV. D) ICGA of a patient demonstrating edges of pigment epithelial detachment (PED) and a polypoidal lesion. E) Optical coherence tomography (OCT) findings of thumb-like polyp (TLP) in peripapillary region (*), PED in parafoveal region and double-layer sign (DLS) located between PED and TLP. F) OCT finding (left) of choriocapillaris segmentation (double-red lines), branching vascular network (gray arrow) and the polypoidal lesions (white arrow), and classic CNV (black arrow) with t the corresponding lesions seen on OCT-Angiography (right). #Images modified from Liu, et al⁶¹, with permission from Wolters-Kluwer Health, Inc. *Images modified from Srour, et al⁷¹, with permission from BMJ Publishing Group.

PATHOPHYSIOLOGY

Histopathology

The histopathological hallmarks of PCV have been described by numerous groups.(15,30–35) These findings include abnormal dilatation of choroidal vessels with hyalinization, massive exudation of fibrin and plasma, arteriosclerosis, and loss of smooth muscle.(15,30–35) One study found choroidal vessels in PCV patients to have areas without CD34 staining, which is indicative of vascular endothelium discontinuity, and positive staining for phosphotungstic acid hematoxylin (PTAH), which is indicative of fibrin.(32) In most cases, RPE had been disrupted, and even when it was preserved, there was choriocapillaris loss.(15,30–35)

Growth Factors and Cytokines

Vascular endothelial growth factor (VEGF) is a molecule that stimulates new vessel growth, increases vascular permeability, and plays a major role in choroidal neovascularization in PCV and NV-AMD.(34,36–41) Multiple studies have found elevated aqueous VEGF levels in PCV patients compared to controls.(41–47) There have been mixed findings when comparing the VEGF levels in PCV patients and NV-AMD patients; some show similar levels, while others show lower levels in the PCV patients.(41–45) Sasaki et al. also found elevated IL-23 as a disease marker (OR = 16.3) for PCV.(44) IL-23 is a pro-angiogenic cytokine, promotes inflammatory responses and is essential for recruitment and activation of inflammatory cells.(44) Other inflammatory cytokines that have been found to be elevated in PCV include, tumor necrosis factor-alpha (45) and monocyte chemotactic protein-1.(44) IL-2, which maintains the activities of regulatory T-cells and mediates immune reactions, was lower in refractory PCV patients compared to stable PCV and control patients.(44)

Genetic Associations of PCV

Several genome-wide association studies (GWAS) of PCV have given us further insight to its pathophysiology and the potential of predicting treatment response. Ma et al. performed a meta-analysis on genetic associations of PCV and found susceptibility single nucleotide polymorphisms (SNPs) in genes in multiple pathways, most notably complement factor H (CFH), age-related maculopathy susceptibility 2 (ARMS2), high-temperature requirement factor A1 (HTRA1), and cholesteryl ester transfer protein (CTEP).(48) CFH mutations are associated with thicker choroid in PCV patients(49) while ARMS2 and HTRA1 mutations are associated with worse PCV disease manifestations and outcomes.(50,51)

CFH protein is part of the inflammatory pathway and is an inhibitor of the alternative complement cascade. Other complement cascade genes that are associated with PCV include C2 and CFB.(48) ARMS2 is expressed in the mitochondria of the outer segment of photoreceptors and its mutation is thought to lead to RPE dysfunction.(52) HTRA1 regulates transforming growth factor beta and matrix metalloproteinases in chronic inflammation and increased HRTA1 induced features of PCV in transgenic mouse models.(53) CTEP protein is a component of reverse cholesterol transport, facilitates transfer of triglycerides, and regulates the concentration of high-density lipoprotein species.(54)

GWAS has also shown certain variants of SNPs influence prognostic features of PCV. ARMS2 LOC387715 rs10490924 have been associated with larger lesion size, higher likelihood of vitreous hemorrhage, and worse visual prognosis 1 year after PDT or combination therapy in PCV patients.(50) CFH 162V variant is associated with choroidal thickness in PCV patients and may be associated with choroidal inflammation in PCV.(48) Park et al. showed risk genotypes, TT of rs10490924 and AA of rs11200638 at ARMS2/HTRA1 were associated with significantly poorer visual outcomes at 1 year follow-up in Korean PCV patients treated with combination therapy of PDT and intravitreal bevacizumab.(51) Those genotypes were associated with less polyp regression, less resolution of leakage on angiography, and worse visual acuity at follow-up.(51)

IMAGING OF PCV

Funduscopic findings alone are often insufficient to diagnose, PCV due to the overlapping features with NV-AMD. Therefore, multiple imaging modalities are often used to differentiate these two entities.

FA and ICGA

Fluorescein angiography (FA) in PCV often shows “occult” choroidal neovascularization (CNV) due to BVNs located in the Bruch’s membrane but one can also see “classic” CNV hyperfluorescent leakage (Figure 1B) due to overlying RPE atrophy or subretinal fibrin deposits.(15) Findings of PCV in FA can mimic NV-AMD and indocyanine green angiography (ICGA) is required to accurately diagnose PCV.(55) ICGA in PCV patients demonstrates abnormal subretinal vascular network (Figure 1C), with aneurysmal dilations or polyp-like lesions (Figure 1D) that are seen often with associated BVNs.(56) Additionally, Kim et. al. reported choroidal vascular hyperpermeability in 43% and punctate hyperfluorescent spots on 53% of PCV patients.(57) They hypothesized these findings were due to leakage associated with punctate hyperpermeable inner choroid spots or late staining of forme fruste drusen or drusen-like subretinal pigment epithelium deposits.(57,58) ICGA is critical in the diagnosis of PCV but is not routinely performed in the initial evaluation of choroidal neovascularization in the United States. Therefore, clinicians must be aware of characteristic signs of PCV on alternative imaging modalities.

OCT

OCT has become an important ancillary test for aiding the diagnosis of PCV. Sato et al. described a double-layer sign (DLS, Figure 1E), which consists of RPE and highly reflective layer underneath, which correlates with BVN on ICGA.(59) Another sign is thumb-like polyp (TLP, Figure 1E), or sharp PED peaks which are described as RPE protrusions on OCT that correspond to polypoidal lesions on ICGA.(59) De Salvo et al carried out a retrospective study to describe OCT findings of PCV complexes on OCT and found multiple PEDs, sharp PED peaks, PED notches, and round polyp lumens to be characteristic findings of PCV.(60) Using these characteristic findings, Liu et al. performed a prospective clinical trial to evaluate spectral-domain OCT (SD-OCT) in differentiating PCV and NV-AMD, and reported that presence of two of the three signs of local PED, DLS, TLP were 89% sensitive and 85% specific for distinguishing PCV from NV-AMD.(61) These studies suggest that OCT can be a useful modality to aid in the diagnosis of PCV.

In addition to imaging the retina, OCT has been used to study the choroidal morphology, using enhanced depth imaging (EDI). Lee et al. used OCT and ICGA to show PCV eyes with normal or subnormal subfoveal choroidal thickness exhibit extrafoveal choroidal thickening at the site of neovascular growth and polypoidal lesions, giving further insight to PCV as a pachychoroidal disease.(62) In 2016, Saito et al. used OCT to monitor response to intravitreal aflibercept injections in PCV patients who were refractory to ranibizumab, and showed significant decrease in subfoveal choroidal thickness and significant improvement in visual acuity.(63) Kim et al. similarly used mean choroidal thicknesses measured by OCT as a treatment outcome after initiation of either intravitreal ranibizumab and aflibercept in three disease groups of NV-AMD, PCV and retinal angiomatous proliferation (RAP) and showed significant differences in amount of decrease in the two treatment groups as well as difference between disease groups.(64) Though OCT measurements of choroidal thickness provide valuable clinical information, some caution using it as a sole measure of treatment response, since it can be influenced by diurnal variations and systemic conditions.(65,66)

En face OCT

En face OCT is a non-invasive imaging modality that is readily available through modification of software viewing option in most SD-OCT machines. En face OCT shows frontal sections of retinal layers providing topographical analysis and assessment of lesion extent. Characteristic findings of PCV include polypoidal lesions that are dome-shaped or round structures confined deeper than the retinal pigment epithelium layer, abnormal choroidal network, and dilated choroidal vessels.(67) Kokame et al. used en face OCT images in a series of eyes diagnosed with PCV by ICGA and found PCV complexes were equally well-visualized in both ICGA and en face OCT, and the extent of PCV complexes were larger using OCT than with ICGA, suggesting that en face OCT can be an effective noninvasive imaging modality for diagnosis of PCV, when ICGA is not available.(68) Alasil et al. also described en face swept source (SS)-OCT features seen in PCV patients including large PEDs with adjoining small PEDs, that often corresponded to the polypoidal lesions seen on ICGA.(69) Although there is currently no consensus on en face OCT criteria for the diagnosis of PCV, further studies may be useful in clinical settings where ICGA is not readily available.

OCT Angiography

OCT angiography (OCT-A) is a relatively new imaging modality that captures retinal and choroidal blood flows by using motion contrasts between OCT images, and aides in the diagnosis of PCV. Unlike ICGA, it is non-invasive, requires no injection of contrast, and has rapid acquisition time compared to traditional dye angiography. Inoue et al. used OCT-A in three cases of PCV and found polypoidal lesions confined between RPE and Bruch’s membrane, and the flow signals within focal regions of the polyps and significant portions of the polyp lumens to be devoid of flow signal.(70) Srour et al. reported BVNs were clearly and consistently detected as hyperflow lesion on OCT-A, but polypoidal lesions had variable signs including hyper-flow round structures surrounded by a hypo-intense halo in some cases and hypo-flow round lesions in most cases (Figure 1F).(71) Hypo-intense round structures are thought to be due to low flow signal within polypoidal lesions.(71) Most studies found a higher detection of PCV polyps using ICGA compared to OCT-A. However, a study by Wang et al. found that using en face OCT-A, BVN lesions were seen more clearly using OCT-A than ICGA.(72) Although OCT-A is currently not a replacement of ICGA, the continued development of the imaging software holds promise for improved detection of PCV lesions using OCT-A.(70,71,73,74)

Fundus Autofluorescence

Fundus autofluorescence (FAF) is another useful non-invasive imaging modality that aids the diagnosis and assessment of treatment response of PCV patients. Ozkok et al. analyzed patients with choroidal neovascularization and found peripapillary fundus autofluorescence defects as an imaging finding that is more associated with PCV compared to typical NV-AMD and RAP.(75) In terms of treatment response, Yamagishi et al. showed elimination of hypoautofluorescent ring in FAF were observed more often in resolved polyps after treatment than with persistent polyps, suggesting FAF’s role in following efficacy of treatment of PCV.(76)

MANAGEMENT OF PCV

The optimal management for PCV is still being investigated. The natural course of PCV remains variable with 50% of patients having a favorable course, in which spontaneous regression of polyps occurs without treatment.(77) In the remaining 50%, there is repeat bleeding, and leakage resulting in RPE and photoreceptor degeneration, scarring and irreversible visual loss. The important endpoints in the treatment of PCV are visual improvement and closure of polypoidal lesions. The predominant options for medical treatment include photodynamic therapy (PDT), anti-VEGF therapy, and combination therapy. Massive submacular hemorrhage in PCV patients may benefit from surgical intervention. Other agents that have been studied in small series/case reports and may have some degree of efficacy include oral steroids, oral eplerenone, and stereotactic radiotherapy; all of these agents require further study before determining their role in the management of PCV.(78–80)

Photodynamic monotherapy

Before the advent of anti-VEGF therapy, PDT was widely used, and remains the main form of treatment for eyes with PCV in Asian countries. The theory behind PDT is that the non-thermal laser reduces perfusion to PCV lesions by inducing local transient occlusion within choroidal vasculature. This, in turn, causes thrombosis of vessels that supply the PCV complexes, and eventually leads to polypoidal regression.(81) It has been demonstrated to induce closure of vascular lesions, regression of polyps, and stabilize visual acuity.(82,83) Although PDT demonstrates favorable results with regression of polyps, there is little regression of the surrounding BVN.(82)

PDT requires less retreatment compared to anti-VEGF therapy.(84,85) Younger age, smaller lesion size, better baseline vision, and less baseline hemorrhage, are predictive measures of favorable one year visual outcomes following PDT monotherapy.(86) Numerous studies have evaluated PDT monotherapy for short and mid-term results and concluded that visual acuity remains stable or improved, while polyp regression occurs in 80–95% of treated eyes.(81,86–95) However, in a meta-analysis performed by Wong et al. on the long-term visual outcome of PCV treated PDT, visual acuity deteriorated at the 3-year mark.(85) Two prospective studies (3- and 5-year follow-up) have demonstrated similar results.(96,97) Adverse events with PDT include sub-retinal hemorrhage, choroidal ischemia, RPE tears and microrips at the margin of the PED, and massive-suprachoroidal hemorrhage.(19,29,98)

Intravitreal anti-VEGF monotherapy

Intravitreal anti-VEGF therapy has demonstrated efficacy in improving neovascular lesions and associated fluid and hemorrhage in NV-AMD. Although anti-VEGF treatments have comparable effects in PCV, there are poorer visual and anatomic effects compared to NV-AMD.(99–101) Anti-VEGF treatments are more commonly utilized in the United States and Europe compared to Asia. When comparing the efficacy of anti-VEGF (ranibizumab) monotherapy to PDT monotherapy, one study found superior visual improvement using ranibizumab (30%) compared to PDT(17%).(102) This finding was also noted in a subsequent study.(82) Ranibizumab, bevacizumab, aflibercept, ziv-aflibercept, and conbercept have all been used as therapeutic anti-VEGF agents for PCV and have been shown to reduce exudation and stabilize or improve vision, but have variable effects on polypoidal lesions and choroidal vascular abnormalities.(55,63,85,102–125)

Only one prospective study has compared anti-VEGF agents directly (bevacizumab and ranibizumab) and found no difference in the number of injections, improvement in vision or decrease in mean central foveal thickness.(111) Retrospective studies comparing aflibercept and ranibizumab have demonstrated similar significant improvements in vision and decreased central foveal thickness for both treatments.(113,115,117,121) However, aflibercept treated eyes had more frequent polyp regression (34–75%) than ranibizumab treated eyes (22%).(113,115,117,121) Additionally, retrospective studies of ranibizumab “non-responders” who were switched to aflibercept treatment demonstrated reduced exudation, reduced choroidal thickness, and stable or improved vision.(63,114) Eyes that were switched from ranibizumab to aflibercept were also noted to have further resolution of PEDs and polyp closure.(117) The differences in outcomes can be due to active transport of aflibercept to the sub-RPE space by RPE cells, compared to gradient diffusion between RPE cells for ranibizumab.(120) Additionally, ranibizumab and bevacizumab only targets VEGF-A, while aflibercept targets VEGF-A, VEGF-B and placental growth factor (PIGF).

Combination therapy

Combination PDT and anti-VEGF treatment theoretically would provide the superior polyp regression effect of PDT with the superior visual outcomes of anti-VEGF therapy. Additionally, anti-VEGF treatments can address the up-regulation of VEGF after PDT and potential subsequent development of secondary CNVs as well as recurrence of PCV.(126–130)

EVEREST was a landmark phase 3, double-blind, multi-center, randomized control clinical trial that compared PDT alone, monthly intravitreal ranibizumab alone, and the combination of PDT with ranibizumab therapy over a 6-month time period. The study showed that complete polyp regression was statistically significantly higher in the combination arm (77.8%) and PDT monotherapy group (71.4%) than with anti-VEGF alone (28.6%).(82) Although the study was not powered to detect differences in visual acuity, there were more letters gained in the combination arm (10.9 ± 10.9 letters) and the ranibizumab monotherapy arm (9.2 ± 12.4 letters), than the PDT monotherapy arm (7.5 ± 10.7 letters). Additional studies with longer follow-up (1 to 3 years) observed similar rates of polyp regression and superior visual outcomes with combination PDT and ranibizumab compared to PDT monotherapy and ranibizumab monotherapy.(131,132) A meta-analysis on combination therapy compared to PDT monotherapy demonstrated no significant difference at three and six months, but the combination cohort had better visual acuity at 12 and 24 months.(84) There was no significant difference in polypoid complex regression, recurrence, decrease in central foveal thickness, or resolution of PED between the two groups.(84)

The EVEREST 2 study is a phase 4, 24-month follow-up, prospective study comparing the effect of ranibizumab monotherapy versus combination ranibizumab and PDT in PCV patients.(133) The PLANET study is a phase 3b/4 prospective study of aflibercept monotherapy compared to combination aflibercept and PDT in PCV patients.(134) These studies will provide additional data to help guide the medical management of PCV.

Intravitreal steroid therapy

A study in PCV patients that compared combined PDT and intravitreal triamcinolone to PDT monotherapy, found no significant difference in mean visual acuity at 2 years between the two groups.(135) Another retrospective study which compared PDT monotherapy to combination PDT and intravitreal bevacizumab and triamcinolone acetonide, found better visual acuity outcomes and longer treatment-free periods with combination therapy.(136) A study by Sakai and colleagues compared outcomes in patients receiving PDT and intravitreal bevacizumab with or without sub-tenon’s kenalog, found a larger percentage of patients with vision improvement and longer treatment-free periods in the group receiving sub-tenon’s kenalog.(137)

Submacular hemorrhage management

Submacular hemorrhage (SMH) is an uncommon, yet serious, complication of PCV that leaves patients with a guarded prognosis. After the initial diagnosis of PCV, the incidence of massive SMH is 2.5% after 1 year and 30% after 10 years.(138) SMH can cause irreversible damage through iron-toxicity, shear stress of photoreceptors by fibrin clots, and separation of photoreceptors from RPE.(139–145) Without treatment, only 11% of eyes have been found to have a BCVA better than 20/200 after 2 years.(146) Longer duration of SMH, height of SMH, extent of hemorrhage, presenting BCVA, and disruption of the ellipsoid zone are all negative predictive factors for visual acuity after SMH in PCV patients.(147–151)

Thin SMHs may be successfully treated by anti-VEGF monotherapy.(152–156) Thicker SMHs may benefit from surgical interventions, such as pneumatic displacement or vitrectomy with pneumatic displacement. For pneumatic displacement, 0.3–0.4mL of an expansile gas is injected through the pars plana, followed by prone positioning.(81) Both intravitreal recombinant tissue plasminogen activator (rTPA) and anti-VEGF agents at the time of pneumatic have been shown to improve visual outcomes.(157) Of note, aflibercept becomes cleaved by rTPA, reducing its overall activity.(158)

Another option for displacement of the subretinal hemorrhage is a pars plana vitrectomy with injection of rTPA in the subretinal space with or without subretinal air.(159) A partial gas fill and prone positioning are necessary to displace the SMH. Intravitreal anti-VEGF injection can also be performed at the time of surgery. External drainage of the subretinal hemorrhage though scleral tunnels has also been described.(61) The management of SMHs is dependent of several factors, including timing, visual prognosis, and general health of the patient.

CONCLUSION

Significant progress has been made in studying the pathophysiology of PCV, the ability of newer imaging modalities to diagnose PCV, and the management of PCV. Recent GWAS identified several genes/loci associated with susceptibility and visual prognosis of PCV patients. Certain growth factors and cytokines have been found to be altered in PCV patients. Advancements in non-invasive imaging modalities, such as OCT, en face OCT, OCT-A has provided clinicians additional tools to diagnose criteria and monitor treatment response in PCV. Current studies support a treatment strategy utilizing a combination of PDT and anti-VEGF.