It has been more than 30 years since polypoidal choroidal vasculopathy (PCV) was first reported in the literature in 1990.1 Until now, the ambiguity of the disease, which was initially described as posterior uveal bleeding syndrome2 or multiple recurrent serosanguineous retinal pigment epithelial (RPE) detachments,3 has been clarified in many ways. PCV has recently been recognized as a variation of neovascular age-related macular degeneration (nAMD) which is more prevalent in Asians than Whites.4 It was included in the newly proposed classification system for age-related macular degeneration (AMD) which encompassed the early phase, in which there is accumulation of different types of extracellular deposits, and the late phase, in which the deposits progress or regress into macular neovascularization (MNV) or macular atrophy (MA).5 In this system, PCV is a variation of type 1 MNV which progresses from pachydrusen.6 Recent advancements in ophthalmic imaging, genetics, and treatment, including information technology, have greatly enhanced our understanding in PCV, which is now considered as one of the leading macular causes of severe, irreversible vision loss worldwide. The aim of this article is to review and update the risk factors, diagnosis, and treatments of PCV.
UPDATES ON RISK FACTORS AND BIOMARKERS FOR PCV DEVELOPMENT
Systemic Factors
Traditional Factors
According to a recent, large-scale, population-based, cross-sectional, retrospective study in Japanese subjects, the previously known systemic risk factors of race, male sex, and smoking, were confirmed to be associated with PCV.7 Other systemic inflammation factors, such as white blood cells, neutrophils, triglyceride, and neutrophil-to-lymphocyte ratio, were also found in significantly higher levels in PCV, compared with controls.8
Genetic Factors
The first 2 reports on genome-wide association study in AMD identified a single nucleotide polymorphisms (SNP) rs1061170 (p.Tyr402His) in the complement factor H (CFH) gene and 2 additional SNPs rs10490924 in PLEKHA1 and rs11200638 in HTRA1 genes, which were significantly associated with AMD.9,10 These genes which are in the linkage loci of 1q31 and 10q26 confirmed the previous identifications of AMD genes by linkage analyses.11 Several additional genes and loci which could explain disease variability were later identified in AMD with the major contribution from variants in CFH, ARMS2-HTRA1, C2-CFB-SKIV2L, and C3 genes. Genes for early stage,12 bilaterality,13 and progression of AMD14 were also identified for AMD by genome-wide association study; however, this approach has not yet identified specific genes for PCV.
With the candidate gene analysis approach, shared SNPs for both AMD and PCV and some distinct SNPs for only PCV were identified. Candidate genes in the pathways other than complement have also been implicated in both nAMD and PCV, such as the angiogenesis pathway [eg, vascular endothelial growth factor (VEGF), placental growth factor (PGF),15 and ANGPT216], high-density lipoprotein (HDL) metabolic pathway [eg, cholesterol ester transfer protein (CETP)17 and ABCG118], fatty acid biosynthesis pathway (eg, ELOVL419), the phagocytic pathway (eg, P2RX4 and P2RX720), and smoking-related inflammation gene (eg, FPR1).21 A recent study revealed significant association between TIE2 gene in the endothelial and cell-specific angiopoietin-Tie pathway of angiogenesis with nAMD and PCV in multiple East Asian cohorts.22
In a meta-analysis, 31 SNPs in 10 AMD-associated genes/loci (ARMS2, HTRA1, CFH, C2, CFB, RDBP, SKIV2L, CETP, 8p21, and 4q12) were found to be significantly associated with PCV, whereas 12 SNPs at the ARMS2-HTRA1 locus showed significant differences between PCV and nAMD, with the stronger effect sizes for the latter. In another candidate gene analysis,23 34 known AMD loci were studied in PCV among East Asians. Significant associations in PCV were identified in 8 loci: ARMS2-HTRA1, CFH, C2-CFB-SKIV2L, CETP, VEGFA, ADAMTS9-AS2, TGFBR1, and COL4A3. It is likely that both diseases are genetically highly correlated and, according to this study, loci for AMD might be accounted for up to 36% of PCV variations. Another study, however, found weaker effect of the association of SRPK2, SKIV2L, PGF, and ARMS2-HTRA1 and PCV.24 The difference in frequency of the risk alleles of some genes (eg, p.Tyr402His and p.Ile62Val in CFH) between Asians and Whites may be one of the factors that play roles in the genetic difference between the 2 diseases.
In a recent study of phenotypes and genotypes of patients with PCV in Portuguese and Singaporean cohorts, CETP rs3764261 and CFH rs800292 (p.Ile62Val) were found to be significantly associated with PCV in each population, respectively. Other phenotypes, such as the prevalence of drusen, polypoidal area, and branching network area, were significantly different.25 Another recent study found a possible association between certain alleles in the Class II human leukocyte antigen gene and PCV patients with Afro-Caribbean descent, suggesting the roles of the immune response modelling in PCV.26
Some recent candidate gene analyses have focused on the association between PCV and CSC. In a meta-analysis, 6 SNPs in 3 genes were found to be significantly associated with CSC. They were ARMS2 rs10490924, CFH rs800292, CFH rs1061170, CFH rs2284664, CFH rs2284664, and TNFRSF10A rs13278062 which is an SNP on tumor necrosis factor receptor superfamily, member 10a gene. It was this SNP which demonstrated the same trend of effect among nAMD, PCV, and CSC.27 Another study found that COL4A3 rs4276018 and B3GALTL rs9564692, which were previously found to be specific for PCV, were significantly associated with MNV in CSC.28
The advent of the next-generation sequencing platforms enables sequencing the whole exome, which is the coding portion of the genome, to further identify new and rare variants of AMD and PCV. In recent exome sequencing analyses in Asians, a rare p.Lys329Arg variant in the FGD6 gene was found to be significantly associated with PCV but not nAMD, whereas another rare variant p.Gln257Arg in the FGD6 gene was identified for nAMD.29 Whether these 2 variants represent independent haplotypes remains to be elucidated. Another rare variant, c.6196A>G, in the IGFN1 gene, was also found to be significantly associated with PCV, but not nAMD.30
Systemic Biomarkers
There is growing evidence for the role of lipid metabolism in AMD and PCV pathogenesis, although the association of serum lipid levels and PCV has been inconsistent. The higher serum levels of HDL and low-density lipoprotein have been reported to be protective factors for PCV, whereas a higher level of total cholesterol may increase the risk of PCV. The same study showed associations of rs10468017 of the hepatic lipase (LIPC) and rs3764261 of the CETP genes, in the HDL metabolic pathway, and PCV, although the combination effects of the serum lipid levels and SNPs was not investigated.31
There has been a recent systematic review and meta-analysis on the association between plasma C-reactive protein, a nonspecific systemic inflammatory protein, and PCV.32 Increased levels of C-reactive protein were found to be associated with PCV.33 The robustness of the results was confirmed by sensitivity analysis or by adjusting for demographics, lifestyle factors, and comorbidities. Other systemic biomarkers which may play roles in inflammation, such as fractalkine, an inflammatory chemokine that guides immune cell migration, were also found to be increased in patients with PCV, compared to age-matched controls.34 Plasma levels of certain metabolites were found to be increased in both patients with nAMD and PCV. The metabolites, hyodeoxycholic acid and l-tryptophanamide, were found to be differently distributed in patients with nAMD and PCV; however, they were found to be genetically associated with nAMD (CFH rs800292 and HTRA1 rs10490924).35
A study demonstrated increased serum levels of matrix metalloproteinase 2 (MMP2) and MMP9 in PCV but not in AMD.36 MMP9, which is involved in extracellular matrix metabolism, was found to be associated with Asian patients with PCV. Another study reported that MMP9 was not found in White patients with PCV.37
Local Ophthalmic Factors
Pachydrusen
Pachydrusen are a recently proposed entity of extracellular deposits, or drusen, in early AMD. The genetic38 and phenotypic characteristics of pachydrusen appear to be distinct from other types of drusen,39 particularly in their association with a thick choroid.6 As more studies of pachydrusen are published, there seems to be more evidence supporting the proposition that they may be a predictive marker for progression into PCV. Older age and more extensive fundus autofluorescence abnormalities have been shown to be associated with pachydrusen.40
In a 5-year, retrospective cohort study of 632 eyes with intermediate AMD, the cumulative incidence of progression to nAMD [including typical nAMD, PCV, and type 3 MNV (retinal angiomatous proliferation, RAP)] from intermediate AMD in eyes with soft drusen or pachydrusen was 17.8% and 17.0%, respectively. However, the presence of pachydrusen was significantly associated with progression to PCV, and not with typical nAMD, and never progressed to RAP. Thicker subfoveal choroidal thickness (SFCT) also increased the risk of progression to PCV. Older age, preexisting nAMD in the fellow eye, and the presence of subretinal drusenoid deposits (SDD) were significantly associated with the risk of progression to nAMD.41
In a cross-sectional, retrospective study of eyes with treatment-naive nAMD including PCV, according to whether or not the fellow eyes contained soft drusen or SDD, they were classified into the “typical drusen” group and “without typical drusen” group. The patients “without typical drusen” were significantly younger, had thicker SFCT at baseline, and a higher proportion of PCV.42
In a retrospective study which aimed to investigate whether the presence of the different types of drusen or PCV/fibrovascular scar in the fellow eyes of patients with PCV influenced the treatment outcome with combined intravitreal aflibercept injection (IAI) and photodynamic therapy (PDT) after a follow-up period of 2 years. There was no significant difference in terms of visual improvement among the different characteristics of the fellow eyes. Although 58.3% of the patients required retreatment within 2 years, the presence of pachydrusen in the fellow eyes predicted better treatment responses. Eyes with pachydrusen in the fellow eyes were found to have significantly fewer retreatments, including the number of additional IAI and combination therapy, and a mean number of additional IAI, as well as the longer retreatment-free period from the initial combination therapy.38
Fellow Eyes
A 5-year retrospective study evaluated the characteristics of the fellow eyes of patients with unilateral PCV, to determine possible risk factors for those fellow eyes progressing to nAMD. Of the 48 fellow eyes, the development of either nAMD or PCV was observed in 8 eyes (17%). SFCT, irregular RPE elevation, choroidal vascular dilation, choroidal vascular hyperpermeability (CVH), and the presence of branching neovascular network (BNN) in the fellow eyes were significantly correlated with the development of PCV.43
Local Biomarkers
One study investigated the lipid profiles in aqueous humor of patients with PCV. In this study of 16 patients with PCV were compared with 28 control subjects. The mean total lipid concentration in the patients with PCV was higher than in control (185.34 μg/mL vs. 165.05 μg/mL), but the difference between the 2 groups was not significant.44 This mean lipid concentration of the controls was similar to previous report in normal subjects, 16.4 mg/dL.45 A total of 19 potential lipid biomarkers were identified including glycerophospholipids, which accounted for the major proportion of the significant differentially expressed lipids. This study suggests that abnormal lipids expression in aqueous humor of PCV patients might participate in the local inflammation or pro-angiogenesis in PCV eyes.44
Although many recent data seem to suggest that local factors, in addition to systemic factors, play roles in PCV development, more evidence from well-designed studies are required to determine their role.
UPDATES ON DIAGNOSIS OF PCV
Diagnostic Imaging
Indocyanine Green Angiography (ICGA)
ICGA is traditionally considered as the gold-standard investigation for the diagnosis of PCV as it allows better visualization of the choroidal circulation for detecting the polypoidal lesions in PCV compared with fluorescein angiography (FA).46 Both confocal scanning laser ophthalmoscope and flash camera–based ICGA systems have been shown to detect over 80% of the polypoidal lesions, although the BNN might be better visualized using the confocal scanning laser ophthalmoscope system.47 Studies evaluating various imaging modalities will utilize ICGA as the reference standard for evaluating their diagnostic accuracies in PCV.
Several studies have utilized ultrawide-field (UWF) ICGA to evaluate the choroidal vasculature of the entire fundus of PCV patients including the peripheral choroid. With the use of binarized images of UWF FA and ICGA, the total vascular area (TVA) and retinal vascular area were measured on the ICGA and FA images respectively. The choroidal vascular area was then calculated by subtracting retinal vascular area from TVA. Finally, the choroidal vascular density (CVD) was obtained by using a choroidal vascular area divided by TVA for the total fundus and 4 specific regions (macular, near-peripheral, mid-peripheral, and far-peripheral regions). A study found that SFCT of PCV eyes significantly correlated with the CVD of the total fundus as well as the 4 regions. CVD was also found to be positively correlated with the presence of choroidal vascular permeability.48 In another study, UWF ICGA was used to evaluate the vortex vein enlargement and choroidal hyperpermeability in patients with PCV and age-matched controls. All PCV eyes were found to have engorgement of vortex veins in 2 or more quadrants on ICGA and, compared with fellow eyes, PCV affected eyes were found to have significantly more quadrants with extended vortex vein engorgement. The area of choroidal hyperpermeability and greatest linear dimension of the PCV lesion were also significantly associated with extended engorged vortex veins. The findings suggest that outflow congestion might be one of the factors involved in the pathogenesis of PCV and UWF ICGA might further enhance the early diagnosis of PCV.49
Spectral-Domain Optical Coherence Tomography (SD-OCT)
SD-OCT has been increasingly used for the diagnosis of PCV due to its ubiquitous availability and noninvasiveness. The most commonly reported SD-OCT features of PCV include sharp-peaked or M-shaped PED, notched PED, multiple PED, double-layer sign with thick choroid or the presence of pachyvessels.46
Several meta-analyses have been carried out to evaluate the accuracy of SD-OCT in diagnosing PCV. Permadi et al conducted a meta-analysis of 7 studies with 911 eyes and found a pooled sensitivity of 0.91, specificity of 0.88, and area under the receiver operating characteristics curve (AUC) of 0.9550. In another meta-analysis, Jiang and Qi51 included 12 studies with a total of 1348 eyes and found that the SD-OCT had a pooled sensitivity of 0.87 and a pooled specificity of 0.83 in diagnosing PCV, giving an AUC of 0.94.
In view of the increasing use of SD-OCT for diagnosing PCV, the Asia-Pacific Ocular Imaging Society (APOIS) PCV workgroup conducted 2 studies in order to develop and validate a set of the non-ICGA features for diagnosing PCV. In the first report by the workgroup,52 15 panel members selected 9 non-ICGA criteria and graded the presence or absence of each of the features in treatment-naive PCV and non-PCV eyes without using ICGA. Based on the grading exercise, 3 SD-OCT features, sub-RPE ring-like lesion, en face OCT showing complex RPE elevation, and sharp-peaked RPE elevations, were found to have AUC of >0.75 and can be considered as major criteria for diagnosing PCV. By combining all these 3 SD-OCT major criteria, the AUC for diagnosing PCV is 0.90 with a sensitivity of 0.75 and specificity of 0.91. The combination of the 3 major criteria was further validated in a separate independent dataset of 55 PCV and 55 typical nAMD patients from 2 different populations and demonstrated an accuracy of 82%.
In the second APOIS report, the authors further evaluated the validity of the non-ICGA features in eyes which had suboptimal response after treated with loading doses of anti-VEGF therapy.53 It was found that in these eyes with persistent PCV activity, the combination of 3 non-ICGA–based criteria (sharp-peaked PED, sub-RPE ring-like lesion and presence of orange nodule on fundus photography) had good agreement compared with ICGA, with an AUC of 0.85. Moreover, SD-OCT-guided PDT was shown to cover 100% area of the polypoidal lesions and 91% area of the BNN.
Optical Coherence Tomography Angiography (OCTA)
OCTA is a more recently developed technology which allows for detection of erythrocyte motion without the need of an intravascular dye, thereby providing high-resolution images of the retinal and choroidal vasculatures.54,55 A number of studies have evaluated the use of OCTA for diagnosing PCV and there appeared to be a wide range of diagnostic accuracy for diagnosing various features of PCV. This is likely due to variations in OCTA analysis protocols and devices, as well as intrinsic factors of the PCV eyes affecting the detection rates.
A study demonstrated the detection rate of en face OCTA to detect active polypoidal lesions shown in ICGA was 65.6%.56 It was found that higher height rather than size and pulsation of the polypoidal lesions was associated with nonvisualization of the lesions on en face OCTA. Presence of thick subretinal hemorrhage or polypoidal lesions located under retinal veins were other reasons for the nonvisualization. In another study, OCTA was able to detect only 10% of polypoidal lesions, which mostly consisted of smaller lesions.57 Another retrospective study found OCTA had excellent correlation with ICGA in determination of lesion size (r=0.997) and was able to identify polypoidal lesions in 69.6% of eyes with PCV.58
With the use of B-scan OCTA, 94.7% of the polypoidal lesions were detected with an accuracy of 92.6% in terms of the number of polypoidal lesions per eye.59 In another retrospective study assessing eyes with PCV using OCTA prior to combined PDT and anti-VEGF therapy, it was found that at baseline, OCTA detected 100% of the polypoidal lesions and 71% of the BNN.60
The visibility of the polypoidal lesions of PCV in OCTA may be dependent on choroid-to-polyp dye infusion time on ICGA. In a retrospective study of PCV eyes using both ICGA and OCTA, Fukuyama et al found that nonvisualized polypoidal lesions on OCTA had longer ICG dye filling time from the first appearance in the choroidal arteries to the first appearance in the lesions, compared to visualized lesions (5.15±2.30 seconds vs. 2.08±1.08 seconds; P<0.001). This means the visibility of polypoidal lesions on OCTA decreases when blood flow is slower.61
A meta-analysis of OCTA in PCV included 573 eyes from 20 studies and evaluated the pooled sensitivity and specificity of OCTA in detecting polypoidal lesions and BNN.62 The detection rate of BNN (0.86) was higher than that of polypoidal lesions (0.67). In terms of diagnostic ability, the combined AUC, sensitivity, and specificity were 0.87, 0.77, and 0.84, respectively.
Multicolor Imaging
Multicolor imaging is a more recently developed technique which has been studied for the evaluation of PCV.63,64 It combines the individual reflectance images from 3 different wavelengths to form a composite pseudocolor image.63 The main advantage of using multicolor imaging is the inclusion of different imaging wavelengths (infrared, green, and blue), and thus it might have the potential to visualize different retinal layers and choroids. In particular, the infrared wavelength of 815 nm can penetrate deeper into the retina as there is minimal absorption from blood or melanin. The RPE and choroidal layer might therefore be better visualized.
In the use of multicolor imaging in 50 consecutive treatment-naive patients with PCV, this technique was able to detect polypoidal lesions in 49 (98%) of 50 eyes, which appeared as dark green oval-shaped lesions.63 Another study also showed comparable rate of detecting PCV at 96% using multicolor imaging.65 The mean number of polypoidal lesions, mean polypoidal lesion area, and mean BNN area for patients with 4 or more polyps were significantly smaller with multicolor imaging compared with ICGA. In contrast, another study conducted by Tan et al64 showed that multicolor imaging only had a sensitivity of 0.5 for detecting polypoidal lesion and was comparable with the sensitivity of 0.48 for CFP. Therefore, whether multicolor imaging can assist the diagnosis of PCV remains to be seen and warrants further investigations.
The recent attempts to use imaging modalities other than ICGA to detect and monitor PCV are in line with the recent decline in using PDT for treatment of PCV. The main advantage of PDT is its better efficacy for closing the polypoidal lesions, which are needed to be localized by ICGA. The use of ICGA may be in further decline if the new treatment options of anti-VEGF for PCV are proven to be as effective as PDT in closing the polypoidal lesions.
Artificial Intelligence (AI)
Recently, deep learning (DL), a new subset of AI, has been found to surpass the performance of traditional machine learning for detecting DR with a sensitivity and specificity of ≥95% in many studies.66 DL was validated for other retinal diseases, including AMD,67 retinal vein occlusion,68 cytomegaloviral retinitis,69 etc., with high performances.
For AMD, there are many studies on DL for tasks such as classification,70 segmentation,71 and prediction72 in a variety of imaging modalities, such as color fundus images, OCT, OCTA, etc. For PCV, on the contrary, there are much fewer studies, most of them focused on classification task with a similar aim of detecting the disease.
The first publications of DL for detecting PCV from ICGA images made use of a publicly available cloud-based AI platform, which was proved to be practical, feasible, without requirement of costly computer hardware or computer experts, to train AI models. Using a 2-step approach, a study trained the first DL model to differentiate between normal and abnormal ICGA images and the second DL model to differentiate between typical nAMD and PCV, an accuracy for detecting PCV in this study was 0.83.73 Another study on ICGA used UWF images of CSC, PCV, and nAMD for developing and training DL models. The precision (positive predictive value) and recall (sensitivity) was found at ~89%.74 These 2 studies claimed that the performance of AI for detecting PCV from ICGA images was comparable to retinal specialists and better than ophthalmology residents.
FA is generally used more often in clinical practice compared to ICGA, an AI model to detect PCV from FA images would be useful in clinical practice. Tsai et al used FA images of PCV and nAMD with corresponded ICGA images from EVEREST Study in Taiwan and Thailand to develop and test a DL model to detect PCV. An accuracy of 82.8% was found for the differentiation between PCV and nAMD. Additional DL model was developed to highlight heatmap regions for detecting PCV lesions on FA where the agreement of 21.9% with experts was found. The third model was applied for segmentation of PCV lesions on FA and achieved the average dice similarity score of 0.88.75 The main limitation of using images from fundus angiography for AI would be the selection of images from a series of time-dependent series. This can be problematic for standardization in both the development of AI and the application of AI in the real world.
Volumetric OCT images of eyes with either nAMD or PCV were used to develop DL models to differentiate these 2 conditions from OCT images. A study applied labels on single B-scan images extracted from each OCT volume and aggregated the probabilistic predictions of all the B-scans within the volume to derive the final prediction. It was found that volume-based classification was not different from the B-scan–based classification with recalls at 0.9 and AUC at 0.97. The saliency maps projected on en face OCT volumes were also found in good correspondence with fundus angiography images of the corresponded OCT volumes.76 There was no comparison between the models and human experts in this study.
Hwang et al published 2 studies on DL for detecting PCV on SD-OCT images. In the first study, DL was developed for a differentiation between RAP and PCV which showed an accuracy, sensitivity, and specificity of 89.1%, 89.4%, and 88.8%, respectively. When distinguishing between AMD (PCV or RAP) and normal, the proposed model had 99.1% accuracy with sensitivity and specificity at 99.2% and 99.1%, respectively. These results were comparable with retinal specialists.77 In the next study, the model was trained and validated based on SD-OCT images of patients with nAMD, PCV, RAP, and healthy controls. The accuracy of this new model was found at 87.4%, which was higher than 2 retinal specialists, 2 retinal fellows, and 4 residents recruited in the study. The F1-scores for detecting normal, PCV, RAP, and nAMD were 1.0, 0.80, 0.83, and 0.69, which were better than the 2 retinal specialists.78
For studies on AI for detecting PCV on CFP, 2 studies were recently published. Both required either additional OCT images or labelling of OCT markers to be analyzed together with CFP for making diagnosis of PCV on CFP. In the first study, a CFP-based model was found to have accuracy, sensitivity, and specificity of 0.78, 0.77, and 0.78, respectively. When combined with OCT biomarkers, such as the presence of double-layer sign, notched PED, and thumb sign, the accuracy, sensitivity, and specificity of the model increased to 0.84, 0.81, and 0.85, respectively.79,80 In another study, the investigators develop 2 DL models, one for CFP and another for OCT, then combined the 2 models for making prediction on both CFP and OCT images to provide outputs as normal, dry AMD, nAMD, or PCV. This bimodal DL, when compared with the model on OCT only or human experts, provided the best performance in terms of sensitivity, specificity, F1-score, and κ score at the values of 89.4, 94.8, 0.894, and 0.841, respectively.80
Most of these AI models for PCV have not yet been validated in new datasets which are different from the developmental data sets.
UPDATES ON TREATMENT OF PCV
Combination Therapy
With the increasing widespread use of anti-VEGF therapy for nAMD, the use of PDT as monotherapy has declined.47,81 One of the first randomized controlled trials (RCTs) for the treatment of PCV (Fujisan Study) showed that initiating the treatment with intravitreal ranibizumab (IVR) monotherapy followed by PDT provided similar visual and anatomical outcomes as initiating the treatment with combined IVR and PDT at 1 year.82
In the later RCTs (the 2 EVEREST studies),83,84 participants were randomized to receive IVR monotherapy or IVR combined with PDT followed by a pro-re-nata (PRN) regimen after 3 initial monthly loading IVR. The combination arm achieved superior visual gains (8.3 vs. 5.1 ETDRS letter gains at 12 mo) and higher rate of polyp closure (77.8% vs. 34.7%). These clinical outcomes were maintained to month 24 when the mean number of IVR was also significantly reduced (12.5 vs. 8.1).84 While the benefits of combining IVR with PDT are clearly demonstrated, the accessibility, cost and requirement of additional expertise and equipment for delivering PDT remains a hurdle for combination therapy to be widely adopted as first-line therapy.85 Around the same time as the EVEREST II study, the PLANET study compared intravitreal aflibercept (IVA) monotherapy to IVA combined with rescue PDT.86,87 In contrast to the PRN regimen used in EVEREST II, IVA was delivered 8-weekly after 3 initial monthly loading doses. The results demonstrated that monotherapy was noninferior to the combination arm (10.7 vs. 10.8 ETDRS letter gains at 12 mo). However, in the PLANET study only a small proportion of the participants met the criteria for rescue PDT, which resulted in 100% of patients receiving aflibercept monotherapy and 85% of the PDT arm receiving only aflibercept monotherapy, which does affect the interpretation of the results of this study. The rate of polypoidal lesion closure was 38.9% and 44.8% in the monotherapy and combination therapy arms, respectively.
The Atlantic study was a 52-week, double-masked, sham-controlled, phase 4, investigator-initiated RCT in naive symptomatic Caucasian patients with PCV.88 Patients were randomized at week 16 to receive IVA monotherapy or combined with PDT following a treat-and-extend (TNE) regimen. At week 52, no significant difference was detected between the IVA monotherapy and the combination arm for best-corrected visual acuity (BCVA) gain (6.5 vs. 5 ETDRS letters), number of injections (8.5 vs. 8.0), rate of complete polyp occlusion (77% vs. 68%), or presence of fluid (68% vs. 57%). One limitation of this study was that only 22% of the eyes underwent PDT treatment. Hence the benefit of combination could not be definitively elucidated.
Beyond RCTs, several real-world studies have also evaluated the potential benefits of combination therapy. Data from the Fight Retinal Blindness! cohort showed that combination therapy achieved larger visual acuity gains (+16.9 letters vs. +8.2 letters), higher proportion of inactive lesion (85.3% vs. 76.8%), quicker time to inactivity (80.7 vs. 150.4 d) and lower number of injections (4.3 vs. 6.4), compared to anti-VEGF monotherapy.89 Another real-world study based in Singapore also reported larger visual acuity gains (+7.7 vs.+5.4 letters) in eyes treated with combination therapy compared to those treated with anti-VEGF monotherapy.90 A retrospective review of patients with PCV in a single tertiary referral center in Taiwan reported similar changes in VA at 3, 6, and 12 months between patients treated with aflibercept monotherapy and those treated with PDT plus ranibizumab.91
The role of adding PDT to eyes with PCV refractory to anti-VEGF monotherapy was evaluated in a retrospective review of the patients in a British cohort.92 The patients were stratified into initial-PDT group and deferred PDT group. Significant improvement in BCVA was seen in the deferred group but not in the initial-PDT group at 18-month follow-up. However, the initial-PDT group required significantly fewer injections after PDT compared to the deferred group. Therefore, PDT as a rescue therapy may be beneficial in the long-term management of PCV, particularly in eyes that had experienced a significant period of prior exposure to anti-VEGF monotherapy.
There are limited reports of treatment outcomes in PCV beyond 2 years, and most are limited to small retrospective series. Challenges to maintaining longer-term visual stability include under-treatment, recurrence, and development of submacular hemorrhage and atrophy. A retrospective review of PCV eyes post-PDT with 3-year follow-up reported vision returned to baseline at 2 years and dropped below baseline at 3 years. Recurrence rate increased from year 1 through year 3; it was associated with worse visual outcome.93
In another retrospective review of eyes with PCV treated with combined PDT and IVR (n=13) or aflibercept (n=30) and completed 5-year follow-up, BCVA significantly improved at 5-year (P=0.01) with the mean time to recurrence at 28.6±23.1 months (95% CI: 21.5=−35.7; median: 18.0).94
There remain concerns regarding the potential to damage the choriocapillaris and RPE with multiple sessions of PDT.95 In a prospective observational study of treatment-naive eyes with symptomatic PCV without MA at baseline, and followed for 5 years after initial PDT,96 the patients were subsequently treated with a PRN regimen of anti-VEGF therapy and/or PDT. MA was found to develop in 41%. The eyes with MA had significantly lower BCVA at 5 years compared with eyes without MA. Eyes with a thin subfoveal choroid at baseline and eyes with more anti-VEGF retreatments were associated with the development of MA involving the fovea at 5 years after initial PDT.
Anti-VEGF Treatment Regimens
One of the early RCTs for PCV (LAPTOP Study) demonstrated that treatment with IVR in PRN regimen could achieve better visual outcomes than PDT monotherapy.97 However, as with typical nAMD, personalized treatment regimen is gaining popularity in PCV management.81,98 TNE regimen has been evaluated in PCV in several studies. In the second year of the PLANET study, investigators were allowed to extend treatment interval up to 12 weeks as an option.87 Accordingly, the treatment number was reduced to a mean of 4.6 between 12 and 24 months. Subsequently, the ALTAIR study compared the safety and efficacy of TNE based on 2-weekly or 4-weekly adjustments.99 The PCV subgroup (n=90), which comprised almost 40% of the study population, gained 7.5 to 8.2 ETDRS letters at week 52 and 3.7 to 4.9 letters at week 96. The mean number of IVA was 3.4 to 3.5 in the second year, and more than 40% of eyes with PCV were extended to 16-week interval at week 96. A multicenter, prospective, nonrandomized study conducted in Japan comprising 49 eyes with PCV and 45 eyes with typical nAMD, patients received IVA according to a TNE regimen following 3 initial monthly loading doses.100 At the end of 2 years, both groups gained 6.5 ETDRS letters. Over half of the study population achieved retreatment interval of 12 weeks at the end of the study period (67.3% in PCV eyes, 51.1% in typical nAMD eyes).
Several real-world series reported IVA using TNE regimen achieved visual acuity gains of up to 9 letters and nearly 50% of patients reaching 12-weekly treatment intervals after 2 years.101–104
The disease activity criteria used to support retreatment decisions commonly comprise fluid on OCT, visual loss, and presence of hemorrhage or leakage. In the ALTAIR study, in addition to shortening and extending, retreatment interval could be maintained if there was persistent but decreasing subretinal fluid without other disease activity indicators.
Other New Therapeutic Options
Brolucizumab and faricimab have been evaluated in the treatment of nAMD but there have been limited data on their efficacy in PCV. For brolucizumab, all the published studies on PCV are from Japan. An analysis of a subgroup of Japanese patients with PCV from the HAWK study reported comparable anatomic and functional outcomes with 8-weekly or 12-weekly injections of brolucizumab, compared with eyes without PCV.105 This study reported the 76% probability of maintaining injections at 12-week interval after the loading phase through 48 weeks, which was similar to the result of a recently published retrospective case series of brolucizumab for PCV.106
In addition, 2 retrospective, multicentered case series on brolucizumab for MNV including PCV conducted in Japan found that after 3 monthly loading injections, complete regression of polypoidal lesions was found in ~80% (78.9% and 82%) of eyes with PCV.107,108 Another retrospective case series in Japan reported the rate of complete closure of the polypoidal lesions at 93.3% with 3 monthly injections of brolucizumab followed by either fixed 8-week or 12-week injections for 1 year. Both visual and anatomical outcomes were also significantly improved with brolucizumab injections.106
Noninfectious intraocular inflammation (IOI) was found from 6.8% to 22.1% of eyes with nAMD or PCV treated with brolucizumab in recent studies of Japanese eyes. Some of these studies contained only eyes with PCV. All the eyes with IOI had amelioration of the inflammation with steroids treatment, except only 1 eye in which vitrectomy was required to treat submacular hematoma and vitreous hemorrhage. All these eyes had no visual loss due to IOI. There have not yet been publications of prospective studies of brolucizumab for PCV whereas the efficacy of faricimab in PCV will be evaluated in an upcoming multinational, prospective clinical trial (SALWEEN study).109
Factors Related to Treatment Response
General Characteristics
Many early studies on the difference between typical nAMD and PCV demonstrated more favorable treatment outcomes of PCV compared with nAMD. This impression was supported by a recent real-world study on 9-year follow-up of 319 eyes, of which half were nAMD and another half were PCV. PCV eyes showed better favorable treatment responses, compared with AMD (final logMAR BCVA 0.9±0.6 vs. 0.6±0.5; P<0.001), with longer maintenance of visual improvement and later onset of significant visual decline.110
A post-hoc analysis of data from the EVEREST II Study was conducted to evaluate the influence of demographic and imaging factors at baseline and month 3 on visual and anatomical outcomes at month 12. In general, younger age and baseline BCVA were significantly associated with better BCVA gains at month 12; whereas baseline smaller polypoidal lesion area was associated with better BCVA at month 12 for IVR monotherapy. In addition, lower baseline central subfield thickness, the treatment with combination therapy at baseline or month 3, the absence of subretinal fluid at month 3 were significantly associated with fluid-free retina at month 12.111
In a post-hoc analysis from the EVEREST Study, PCV were classified into 3 different subtypes according to the characteristics of polypoidal lesions and BNN.112 Type A PCV, which had polypoidal lesions with interconnecting channels,113 had the best BCVA gain (13 letters from baseline) and the best final BCVA when compared with other subtypes at month 6 (80.1 letters in type A, 67.2 in type B, 50.4 in type C; P<0.001).112
Specific Characteristics
SFCT. SFCT was found to be a factor associated with treatment response in several case series. Thicker SFCT was associated with poorer response to anti-VEGF monotherapy in Asians114 and Whites with a proposed cutoff SFCT value for poor response at 257 µm.115 Another study also found that SFCT was the only independent parameter for response to the 3-monthly anti-VEGF injections (bevacizumab, ranibizumab, or aflibercept) in Asians and proposed another cutoff SFCT value, which might be used to classify PCV as pachychoroid or nonpachychoroid, at 267.5 μm.116
In another series, SFCT was found not to be associated with response to either with IVR or IVA. Thin nasal peripapillary choroidal thickness was found to be significantly associated with fewer number of anti-VEGF injections throughout 12 months. A SFCT/nPCT ratio was proposed as a predictor for treatment response; a high ratio indicating better response.117
Choroidal Vascular Hyperpermeability, Index, and Density. Many parameters in choroidal vasculatures were evaluated as treatment outcomes for PCV. These parameters, including vascular hyperpermeability, vascular index, and vascular density, have different measurement methods but are implicated as disease indicators in choroidal vasculature in PCV.
In a prospective series of PCV treated with combination therapy (31 eyes) or anti-VEGF monotherapy (41 eyes), CVH from ICGA was found to be associated with better visual outcome and lower number of injections in patients received combination therapy. However, choroidal thickness was not found to be associated with BCVA gain or treatment requirement. The presence of CVH was not associated with BCVA gain in the anti-VEGF monotherapy group. This inconclusive result on CVH as an outcome predictor was also found in another study on the treatment of PCV with anti-VEGF.118
CVD identified from UWF ICGA was also found to be higher than controls in most regions of the choroid including periphery in another series of 32 naive eyes with PCV and 30 eyes of normal controls. The higher CVD was also associated with thicker choroid, CVH, and poor response to the anti-VEGF treatments (bevacizumab, ranibizumab, or aflibercept).49
A retrospective case series of PCV with 5-year follow-up focused on choroidal factors related to disease recurrence. Of the 147 eyes, only 26 eyes were without recurrence. Younger age at onset, higher choroidal vascularity index, and presence of CVH were related with low risk of recurrence. For those with recurrence, having large lesion, and cluster polypoidal lesions were found to be associated with worse final visual acuity.119
RPE and Retina. In a retrospective series of patients with PCV treated with either ranibizumab or aflibercept in a PRN regimen and 24-month follow-up, 10.5% developed RPE atrophy.120 Baseline SFCT was significantly thinner and presence of SDD was significantly more frequent in eyes with the RPE atrophy.120 Another retrospective series followed patients with PCV treated with either ranibizumab monotherapy or combined with PDT for 5 years, chorioretinal atrophy (CRA) was found in 40.8%.121 The absence of subretinal fluid, the presence of intraretinal fluid, thin choroid, and history of PDT were significant risk factors for the development of CRA. Faster growth of the CRA was related to thin choroid and the presence of subretinal hyperreflective material.121
In another retrospective series, PCV eyes were treated with 3 initial monthly injections of IVR followed by PRN regimen for 24 months; 87% of the eyes had final BCVA of 20/40 or better. The absence of baseline PED was significantly associated with requiring no IVR between 12 to 23 months. The presence of serous retinal detachment (SRD) at months 6 and 12 was associated with significantly more retreatment rates from months 12 to 23, when compared to eyes without SRD.122
In another retrospective series of patients who underwent combined therapy of PDT with anti-VEGF therapy (ranibizumab or aflibercept), the prevalence of BNN in OCTA in the first month after the treatment was significantly higher in the disease active group, compared with the inactive group. The active group was characterized by recurrence or persistence of 2 specific findings: SRD or subretinal hemorrhage within 3 months of the treatment.123
Pharmacogenetics. In the first systematic review and meta-analysis of association between genetic variants and anti-VEGF response in PCV, the genetic polymorphisms CFH I62V (rs800292), CFH Y402H (rs1061170), HTRA1-62A/G (rs11200638), and ARMS2 A69S (rs10490924) were found to be related to variable treatment response.124 Although ARMS2 A69S (rs10490924) was previously found to be associated with various phenotypes of PCV, the meta-analysis in this study did not find a significant association with response after anti-VEGF therapy in both dominant and recessive models due to inconsistency in statistical homogeneity in each model.124
In a recent pharmacogenetics study of aflibercept treatment in PCV, patients with nAMD and PCV who received as-needed aflibercept injections, A allele of C2-CFB-SKIV2L rs429608 was related with improved visual outcome at month 12. Both T allele of ARMS2 A69S and C allele of CFH rs1329428 were significantly related with retreatment and additional anti-VEGF injections.125
Summary and Future Directions
There are many areas where future studies on PCV can address unmet needs. One of the important research questions may still rest on the phenotypic manifestation of the polypoidal lesions which is unique for PCV. Why do these lesions occur more frequently in Asians and more commonly in association with thick choroid? Large-scale, well-conducted, population-based studies in Asia with the application of the latest technology in genetic sequencing and imaging may help address this question and elucidate the relationship between PCV and other diseases in pachychoroid disease spectrum. A prospective longitudinal study of pachydrusen may be required to support that this entity is a precursor of PCV. The non-ICGA criteria may be evaluated for real-world monitoring and assisting in treatment of PCV. The studies of DL in PCV may be shifted from diagnosis to disease monitoring and outcome prediction using automated segmentation of OCT images, similar to many studies of DL in nAMD. While insight into PCV may help pave ways for future treatments, applying the generally available anti-VEGF agents in a TNE treatment regimen is maintaining treatment outcomes with lesser burden to patients. Newly approved anti-VEGF agents for nAMD may provide better durability, however, well-designed studies are required to confirm their efficacy for PCV as well as the role for combination therapy of these new anti-VEGF agents, as used in the real world, with PDT.
REFERENCES
1. Yannuzzi LA, Sorenson J, Spaide RF, et al. Idiopathic
polypoidal choroidal vasculopathy (IPCV).
Retina (Philadelphia, Pa). 1990;10:1–8.
2. Kleiner RC, Brucker AJ, Johnston RL. The posterior uveal bleeding syndrome.
Retina (Philadelphia, Pa). 1990;10:9–17.
3. Stern RM, Zakov ZN, Zegarra H, et al. Multiple recurrent serosanguineous retinal pigment epithelial detachments in black women. Am J Ophthalmol. 1985;100:560–569.
4. Wong CW, Yanagi Y, Lee WK, et al.
Age-related macular degeneration and
polypoidal choroidal vasculopathy in Asians. Prog Retin Eye Res. 2016;53:107–139.
5. Spaide RF. Improving the
age-related macular degeneration contruct: a new classification system.
Retina (Philadelphia, Pa). 2018;38:891–899.
6. Spaide RF. Disease expression in nonexudative
age-related macular degeneration varies with choroidal thickness.
Retina (Philadelphia, Pa). 2018;38:708–716.
7. Fujiwara K, Yasuda M, Hata J, et al. Prevalence and risk factors for
polypoidal choroidal vasculopathy in a general Japanese population: the Hisayama study. Semin Ophthalmol. 2018;33:813–819.
8. Pan M, Zhou P, Liu Z, et al. Peripheral complete blood cell count indices and serum lipid levels in
polypoidal choroidal vasculopathy. Clin Exp Optom. 2022:1–6. [Epub ahead of print].
9. Klein RJ, Zeiss C, Chew EY, et al. Complement factor H polymorphism in
age-related macular degeneration. Science (New York, NY). 2005;308:385–389.
10. Dewan A, Liu M, Hartman S, et al. HTRA1 promoter polymorphism in wet
age-related macular degeneration. Science (New York, NY). 2006;314:989–992.
11. Weeks DE, Conley YP, Tsai HJ, et al. Age-related maculopathy: a genomewide scan with continued evidence of susceptibility loci within the 1q31, 10q26, and 17q25 regions. Am J Hum Genet. 2004;75:174–189.
12. Holliday EG, Smith AV, Cornes BK, et al. Insights into the genetic architecture of early stage
age-related macular degeneration: a genome-wide association study meta-analysis. PLoS ONE. 2013;8:e53830.
13. Kawashima-Kumagai K, Yamashiro K, Yoshikawa M, et al. A genome-wide association study identified a novel genetic loci STON1-GTF2A1L/LHCGR/FSHR for bilaterality of neovascular
age-related macular degeneration. Sci Rep. 2017;7:7173.
14. Yan Q, Ding Y, Liu Y, et al. Genome-wide analysis of disease progression in
age-related macular degeneration. Hum Mol Genet. 2018;27:929–940.
15. Chen LJ, Ma L, Chu WK, et al. Identification of PGF as a new gene for neovascular
age-related macular degeneration in a Chinese population. Invest Ophthalmol Vis Sci. 2016;57:1714–1720.
16. Ma L, Brelen ME, Tsujikawa M, et al. Identification of ANGPT2 as a new gene for neovascular
age-related macular degeneration and
polypoidal choroidal vasculopathy in the Chinese and Japanese populations. Invest Ophthalmol Vis Sci. 2017;58:1076–1083.
17. Liu K, Chen LJ, Lai TY, et al. Genes in the high-density lipoprotein metabolic pathway in
age-related macular degeneration and
polypoidal choroidal vasculopathy. Ophthalmology. 2014;121:911–916.
18. Ma L, Liu K, Tsujikawa M, et al. Association of ABCG1 with neovascular
age-related macular degeneration and
polypoidal choroidal vasculopathy in Chinese and Japanese. Invest Ophthalmol Vis Sci. 2016;57:5758–5763.
19. Conley YP, Thalamuthu A, Jakobsdottir J, et al. Candidate gene analysis suggests a role for fatty acid biosynthesis and regulation of the complement system in the etiology of age-related maculopathy. Hum Mol Genet. 2005;14:1991–2002.
20. Gu BJ, Baird PN, Vessey KA, et al. A rare functional haplotype of the P2RX4 and P2RX7 genes leads to loss of innate phagocytosis and confers increased risk of
age-related macular degeneration. FASEB J. 2013;27:1479–1487.
21. Liang XY, Chen LJ, Ng TK, et al. FPR1 interacts with CFH, HTRA1 and smoking in exudative
age-related macular degeneration and
polypoidal choroidal vasculopathy. Eye. 2014;28:1502–1510.
22. Chen ZJ, Ma L, Brelen ME, et al. Identification of TIE2 as a susceptibility gene for neovascular
age-related macular degeneration and
polypoidal choroidal vasculopathy. Br J Ophthalmol. 2021;105:1035–1040.
23. Fan Q, Cheung CMG, Chen LJ, et al. Shared genetic variants for
polypoidal choroidal vasculopathy and typical neovascular
age-related macular degeneration in East Asians. J Hum Genet. 2017;62:1049–1055.
24. Ma L, Li Z, Liu K, et al. Association of genetic variants with
polypoidal choroidal vasculopathy: a systematic review and updated meta-analysis. Ophthalmology. 2015;122:1854–1865.
25. Jordan-Yu JM, Teo K, Fan Q, et al. T and genetic variations between Asian and Caucasian
polypoidal choroidal vasculopathy. Br J Ophthalmol. 2021;105:1716–1723.
26. Merle H, Béral L, Rocher M, et al. Class II human leukocyte antigen (HLA) and susceptibility to
polypoidal choroidal vasculopathy in Afro-Caribbean descent. Clin Ophthalmol. 2022;16:1047–1053.
27. Chen ZJ, Lu SY, Rong SS, et al. Genetic associations of central serous chorioretinopathy: a systematic review and meta-analysis. Br J Ophthalmol. 2021. [Epub ahead of print].
28. Mori Y, Miyake M, Hosoda Y, et al. Genome-wide survival analysis for macular neovascularization development in central serous chorioretinopathy revealed shared genetic susceptibility with
polypoidal choroidal vasculopathy. Ophthalmology. 2022. [Epub ahead of print].
29. Cheng CY, Yamashiro K, Chen LJ, et al. New loci and coding variants confer risk for
age-related macular degeneration in East Asians. Nat Commun. 2015;6:6063.
30. Wen X, Liu Y, Yan Q, et al. Association of IGFN1 variant with
polypoidal choroidal vasculopathy. J Gene Med. 2018;20:e3007.
31. Xu N, Xu H, Zhao M, et al. Associations of systemic, serum lipid and lipoprotein metabolic pathway gene variations with
polypoidal choroidal vasculopathy in China. PLoS ONE. 2019;14:e0226763.
32. Zhou H, Zhao X, Chen Y. Plasma cytokine profiles in patients with
polypoidal choroidal vasculopathy and neovascular
age-related macular degeneration. Asia Pac J Ophthalmol (Phila). 2022;11:536–542.
33. Borgersen NJ, Møller-Lorentzen T, Sørensen TL, et al. Association between C-reactive protein and
polypoidal choroidal vasculopathy: a systematic review and meta-analysis. Acta Ophthalmol (Copenh). 2021;99:470–477.
34. Sørensen J, Subhi Y, Molbech CR, et al. Plasma levels of inflammatory chemokines in patients with
polypoidal choroidal vasculopathy. Acta Ophthalmol (Copenh). 2020;98:384–389.
35. Deng Y, Shuai P, Wang H, et al. Untargeted metabolomics for uncovering plasma biological markers of wet
age-related macular degeneration. Aging. 2021;13:13968–14000.
36. Zeng R, Wen F, Zhang X, et al. Serum levels of matrix metalloproteinase 2 and matrix metalloproteinase 9 elevated in
polypoidal choroidal vasculopathy but not in
age-related macular degeneration. Mol Vis. 2013;19:729–736.
37. Sørensen J, Subhi Y, Molbech CR, et al. Plasma levels of matrix metalloprotease MMP-9 and tissue inhibitor TIMP-1 in Caucasian patients with
polypoidal choroidal vasculopathy. Vision (Basel, Switzerland). 2020;98:384–389.
38. Matsubara M, Sakurada Y, Sugiyama A, et al. Response to photodynamic therapy combined with intravitreal aflibercept for
polypoidal choroidal vasculopathy depending on fellow-eye condition: 2-year results. PLoS ONE. 2020;15:e0237330.
39. Fukuda Y, Sakurada Y, Yoneyama S, et al. Clinical and genetic characteristics of pachydrusen in patients with exudative
age-related macular degeneration. Sci Rep. 2019;9:11906.
40. Lai TYY, Tang Z, Lai ACW, et al. Association of fundus autofluorescence abnormalities and pachydrusen in central serous chorioretinopathy and
polypoidal choroidal vasculopathy. J Clin Med. 2022;11:5340.
41. Kim KL, Joo K, Park SJ, et al. Progression from intermediate to neovascular
age-related macular degeneration according to drusen subtypes: Bundang AMD cohort study report 3. Acta Ophthalmol (Copenh). 2022;100:e710–e718.
42. Kamao H, Goto K, Matsuno K, et al. Clinical characteristics of neovascular
age-related macular degeneration without typical drusen. J Ophthalmol. 2021;2021:6683532.
43. Kim K, Kim JM, Kim DG, et al. Five-year follow-up of unaffected fellow eyes in patients with
polypoidal choroidal vasculopathy. Ophthalmologica. 2020;243:172–177.
44. Liu B, Cong C, Li Z, et al. Analysis of the aqueous humor lipid profile in patients with
polypoidal choroidal vasculopathy. Exp Eye Res. 2022;222:109160.
45. Jahn CE, Leiss O, von Bergmann K. Lipid composition of human aqueous humor. Ophthalmic Res. 1983;15:220–224.
46. Cheung CMG, Lai TYY, Ruamviboonsuk P, et al.
Polypoidal choroidal vasculopathy: definition, pathogenesis, diagnosis, and management. Ophthalmology. 2018;125:708–724.
47. Cheung CM, Lai TY, Chen SJ, et al. Understanding indocyanine green angiography in
polypoidal choroidal vasculopathy: the group experience with digital fundus photography and confocal scanning laser ophthalmoscopy.
Retina (Philadelphia, Pa). 2014;34:2397–2406.
48. Ryu G, Moon C, van Hemert J, et al. Quantitative analysis of choroidal vasculature in
polypoidal choroidal vasculopathy using ultra-widefield indocyanine green angiography. Sci Rep. 2020;10:18272.
49. Jeong A, Lim J, Sagong M. Choroidal vascular abnormalities by ultra-widefield indocyanine green angiography in
polypoidal choroidal vasculopathy. Invest Ophthalmol Vis Sci. 2021;62:29.
50. Permadi AC, Djatikusumo A, Adriono GA. Optical coherence tomography in diagnosing
polypoidal choroidal vasculopathy. Looking into the future: a systematic review and meta-analysis. Int J
Retina Vitreous. 2022;8:14.
51. Jiang Y, Qi S. Diagnostic value of spectral-domain optical coherence tomography for
polypoidal choroidal vasculopathy: a systematic review and meta-analysis. Front Med (Lausanne). 2022;9:878946.
52. Cheung CMG, Lai TYY, Teo K, et al.
Polypoidal choroidal vasculopathy: consensus nomenclature and non-indocyanine green angiograph diagnostic criteria from the Asia-Pacific Ocular Imaging Society PCV workgroup. Ophthalmology. 2021;128:443–452.
53. Chong Teo KY, Sadda SR, Gemmy Cheung CM, et al. Non-ICGA treatment criteria for suboptimal anti-VEGF response for
polypoidal choroidal vasculopathy: APOIS PCV workgroup report 2. Ophthalmol
Retina. 2021;5:945–953.
54. Ng DS, Chan LK, Ng CM, et al. Visualising the choriocapillaris: histology, imaging modalities and clinical research - a review. Clin Exp Ophthalmol. 2022;50:91–103.
55. Bacherini D, Mastropasqua R, Borrelli E, et al. OCT-A in the management of vitreoretinal diseases and surgery. Asia Pac J Ophthalmol (Phila). 2021;10:12–19.
56. Zhan Z, Sun L, Jin C, et al. Comparison between non-visualized polyps and visualized polyps on optical coherence tomography angiography in
polypoidal choroidal vasculopathy. Graefes Arch Clin Exp Ophthalmol. 2019;257:2349–2356.
57. Eriş E, Kocakaya AE. Comparison of optical coherence tomography angiography and green indocyanine angiography in
polypoidal choroidal vasculopathy: a prospective study. J Fr Ophtalmol. 2019;42:690–695.
58. Singh SR, Goyal P, Parameswarappa DC, et al. Angiographic features of
polypoidal choroidal vasculopathy using indocyanine green angiography and optical coherence tomography angiography: a comparative study. Eur J Ophthalmol. 2020;30:1076–1081.
59. Fujita A, Kataoka K, Takeuchi J, et al. Diagnostic characteristics of
polypoidal choroidal vasculopathy based on B-scan swept-source optical coherence tomography angiography and its interrater agreement compared with indocyanine green angiography.
Retina (Philadelphia, Pa). 2020;40:2296–2303.
60. Wang WY, Yang CH, Chen TC, et al. Quantitative analysis of branching neovascular networks in
polypoidal choroidal vasculopathy by optical coherence tomography angiography after photodynamic therapy and anti-vascular endothelial growth factor combination therapy. Graefes Arch Clin Exp Ophthalmol. 2022;260:2249–2260.
61. Fukuyama H, Iwami H, Araki T, et al. Indocyanine green dye filling time for polypoidal lesions in
polypoidal choroidal vasculopathy affects the visibility of the lesions on OCT angiography. Ophthalmol
Retina. 2018;2:803–807.
62. Wang Y, Yang J, Li B, et al. Detection rate and diagnostic value of optical coherence tomography angiography in the diagnosis of
polypoidal choroidal vasculopathy: a systematic review and meta-analysis. J Ophthalmol. 2019;2019:6837601.
63. Tan CS, Ting DS, Lim LW. Multicolor fundus imaging of
polypoidal choroidal vasculopathy. Ophthalmol
Retina. 2019;3:400–409.
64. Tan ACS, Yanagi Y, Cheung GCM. Comparison of multicolor imaging and color fundus photography in the detection of pathological findings in eyes with
polypoidal choroidal vasculopathy.
Retina (Philadelphia, Pa). 2020;40:1512–1519.
65. Lim LW, Tan CS, Ting DS. Comparison of
polypoidal choroidal vasculopathy lesion sizes measured on multicolor imaging and indocyanine green angiography. Transl Vis Sci Technol. 2021;10:35.
66. Ruamviboonsuk P, Krause J, Chotcomwongse P, et al. Erratum: author correction: deep learning versus human graders for classifying diabetic retinopathy severity in a nationwide screening program. NPJ digital medicine. 2019;2:68.
67. Dow ER, Keenan TDL, Lad EM, et al. From data to deployment: the Collaborative Community on Ophthalmic Imaging roadmap for artificial intelligence in
age-related macular degeneration. Ophthalmology. 2022;129:e43–e59.
68. Yeung L, Lee YC, Lin YT, et al. Macular ischemia quantification using deep-learning denoised optical coherence tomography angiography in branch retinal vein occlusion. Transl Vis Sci Technol. 2021;10:23.
69. Srisuriyajan P, Cheewaruangroj N, Polpinit P, et al. Cytomegalovirus retinitis screening using machine learning technology.
Retina (Philadelphia, Pa). 2022;42:1709–1715.
70. Burlina PM, Joshi N, Pacheco KD, et al. Use of deep learning for detailed severity characterization and estimation of 5-year risk among patients with
age-related macular degeneration. JAMA Ophthalmol. 2018;136:1359–1366.
71. Moraes G, Fu DJ, Wilson M, et al. Quantitative analysis of OCT for neovascular
age-related macular degeneration using deep learning. Ophthalmology. 2021;128:693–705.
72. Banerjee I, de Sisternes L, Hallak JA, et al. Prediction of
age-related macular degeneration disease using a sequential deep learning approach on longitudinal SD-OCT imaging biomarkers. Sci Rep. 2020;10:15434.
73. Yang J, Zhang C, Wang E, et al. Utility of a public-available artificial intelligence in diagnosis of
polypoidal choroidal vasculopathy. Graefes Arch Clin Exp Ophthalmol. 2020;258:17–21.
74. Kim IK, Lee K, Park JH, et al. Classification of pachychoroid disease on ultrawide-field indocyanine green angiography using auto-machine learning platform. Br J Ophthalmol. 2021;105:856–861.
75. Tsai YY, Lin WY, Chen SJ, et al. Diagnosis of
polypoidal choroidal vasculopathy from fluorescein angiography using deep learning. Transl Vis Sci Technol. 2022;11:6.
76. Ma D, Kumar M, Khetan V, et al. Clinical explainable differential diagnosis of
polypoidal choroidal vasculopathy and
age-related macular degeneration using deep learning. Comput Biol Med. 2022;143:105319.
77. Hwang DD, Choi S, Ko J, et al. Distinguishing retinal angiomatous proliferation from
polypoidal choroidal vasculopathy with a deep neural network based on optical coherence tomography. Sci Rep. 2021;11:9275.
78. Han J, Choi S, Park JI, et al. Classifying neovascular
age-related macular degeneration with a deep convolutional neural network based on optical coherence tomography images. Sci Rep. 2022;12:2232.
79. Chou YB, Hsu CH, Chen WS, et al. Deep learning and ensemble stacking technique for differentiating
polypoidal choroidal vasculopathy from neovascular
age-related macular degeneration. Sci Rep. 2021;11:7130.
80. Xu Z, Wang W, Yang J, et al. Automated diagnoses of
age-related macular degeneration and
polypoidal choroidal vasculopathy using bi-modal deep convolutional neural networks. Br J Ophthalmol. 2021;105:561–566.
81. Fenner BJ, Cheung CMG, Sim SS, et al. Evolving treatment paradigms for PCV. Eye. 2022;36:257–265.
82. Gomi F, Oshima Y, Mori R, et al. Initial versus delayed photodynamic therapy in combination with ranibizumab for treatment of
polypoidal choroidal vasculopathy: the Fujisan study.
Retina (Philadelphia, Pa). 2015;35:1569–1576.
83. Koh A, Lai TYY, Takahashi K, et al. Efficacy and safety of ranibizumab with or without verteporfin photodynamic therapy for
polypoidal choroidal vasculopathy: a randomized clinical trial. JAMA Ophthalmol. 2017;135:1206–1213.
84. Lim TH, Lai TYY, Takahashi K, et al. Comparison of ranibizumab with or without verteporfin photodynamic therapy for
polypoidal choroidal vasculopathy: the EVEREST II randomized clinical trial. JAMA Ophthalmol. 2020;138:935–942.
85. Sirks MJ, van Dijk EHC, Rosenberg N, et al. Clinical impact of the worldwide shortage of verteporfin (Visudyne®) on ophthalmic care. Acta Ophthalmol (Copenh). 2022. [Epub ahead of print].
86. Lee WK, Iida T, Ogura Y, et al. Efficacy and safety of intravitreal aflibercept for
polypoidal choroidal vasculopathy in the PLANET study: a randomized clinical trial. JAMA Ophthalmol. 2018;136:786–793.
87. Wong TY, Ogura Y, Lee WK, et al. Efficacy and safety of intravitreal aflibercept for
polypoidal choroidal vasculopathy: two-year results of the aflibercept in
polypoidal choroidal vasculopathy study. Am J Ophthalmol. 2019;204:80–89.
88. Silva R, Arias L, Nunes S, et al. Efficacy and safety of intravitreal aflibercept treat and extend for
polypoidal choroidal vasculopathy in the ATLANTIC Study: a randomized clinical trial. Ophthalmologica. 2022;245:80–90.
89. Chong Teo KY, Squirrell DM, Nguyen V, et al. A multicountry comparison of real-world management and outcomes of
polypoidal choroidal vasculopathy: Fight Retinal
Blindness! Cohort. Ophthalmol
Retina. 2019;3:220–229.
90. Fenner BJ, Ting DSW, Tan ACS, et al. Real-world treatment outcomes of
age-related macular degeneration and
polypoidal choroidal vasculopathy in Asians. Ophthalmol
Retina. 2020;4:403–414.
91. Ma IH, Hsia Y, Hsieh YT, et al. Real world experience of the treatment outcome between photodynamic therapy combined with ranibizumab and aflibercept monotherapy in
polypoidal choroidal vasculopathy. Sci Rep. 2021;11:20115.
92. Liu S, Chhabra R. Real-world outcomes of combined therapy of photodynamic therapy with anti-vascular endothelial growth factor for
polypoidal choroidal vasculopathy. Eye. 2022;36:1934–1939.
93. Wong CW, Cheung CM, Mathur R, et al. Three-year results of
polypoidal choroidal vasculopathy treated with photodynamic therapy: retrospective study and systematic review.
Retina (Philadelphia, Pa). 2015;35:1577–1593.
94. Wataru K, Sugiyama A, Yoneyama S, et al. Five-year outcomes of photodynamic therapy combined with intravitreal injection of ranibizumab or aflibercept for
polypoidal choroidal vasculopathy. PLoS ONE. 2020;15:e0229231.
95. Teo KYC, Yanagi Y, Lee SY, et al. Comparison of optical coherence tomography angiographic changes after anti-vascular endothelial growth factor therapy alone or in combination with photodynamic therapy in
polypoidal choroidal vasculopathy.
Retina (Philadelphia, Pa). 2018;38:1675–1687.
96. Kawai K, Miyata M, Ooto S, et al. Macular atrophy at 5 years after photodynamic therapy for
polypoidal choroidal vasculopathy. Eye. 2022. [Epub ahead of print].
97. Oishi A, Kojima H, Mandai M, et al. Comparison of the effect of ranibizumab and verteporfin for
polypoidal choroidal vasculopathy: 12-month LAPTOP study results. Am J Ophthalmol. 2013;156:644–651.
98. Chaikitmongkol V, Sagong M, Lai TYY, et al. Treat-and-extend regimens for the management of neovascular
age-related macular degeneration and
polypoidal choroidal vasculopathy: consensus and recommendations from the Asia-Pacific Vitreo-
retina Society. Asia Pac J Ophthalmol (Phila). 2021;10:507–518.
99. Okada AA, Takahashi K, Ohji M, et al. Efficacy and safety of intravitreal aflibercept treat-and-extend regimens in the ALTAIR study: 96-week outcomes in the
polypoidal choroidal vasculopathy subgroup. Adv Ther. 2022;39:2984–2998.
100. Maruko I, Ogasawara M, Yamamoto A, et al. Two-year outcomes of treat-and-extend intravitreal aflibercept for exudative
age-related macular degeneration: a prospective study. Ophthalmol
Retina. 2020;4:767–776.
101. Hosokawa M, Morizane Y, Hirano M, et al. One-year outcomes of a treat-and-extend regimen of intravitreal aflibercept for
polypoidal choroidal vasculopathy. Jpn J Ophthalmol. 2017;61:150–158.
102. Tamachi T, Kohno T, Yamamoto M, et al. One-year results of a treat-and-extend regimen of intravitreal aflibercept for
polypoidal choroidal vasculopathy. Ophthalmol Ther. 2020;9:1069–1082.
103. Morimoto M, Matsumoto H, Mimura K, et al. Two-year results of a treat-and-extend regimen with aflibercept for
polypoidal choroidal vasculopathy. Graefes Arch Clin Exp Ophthalmol. 2017;255:1891–1897.
104. Morizane-Hosokawa M, Morizane Y, Kimura S, et al. Impact of polyp regression on 2-year outcomes of intravitreal aflibercept injections: a treat-and-extend regimen for
polypoidal choroidal vasculopathy. Acta Med Okayama. 2018;72:379–385.
105. Ogura Y, Jaffe GJ, Cheung CMG, et al. Efficacy and safety of brolucizumab versus aflibercept in eyes with
polypoidal choroidal vasculopathy in Japanese participants of HAWK. Br J Ophthalmol. 2022;106:994–999.
106. Ito A, Maruyama-Inoue M, Kitajima Y, et al. One-year outcomes of intravitreal brolucizumab injections in patients with
polypoidal choroidal vasculopathy. Sci Rep. 2022;12:7987.
107. Matsumoto H, Hoshino J, Mukai R, et al. Short-term outcomes of intravitreal brolucizumab for treatment-naïve neovascular
age-related macular degeneration with type 1 choroidal neovascularization including
polypoidal choroidal vasculopathy. Sci Rep. 2021;11:6759.
108. Tanaka K, Koizumi H, Tamashiro T, et al. Short-term results for brolucizumab in treatment-naïve neovascular
age-related macular degeneration: a Japanese multicenter study. Jpn J Ophthalmol. 2022;66:379–385.
109. A study evaluating the effectiveness and safety of faricimab (RO6867461) in participants with
polypoidal choroidal vasculopathy. 2022. January 12, 2023.
https://www.isrctn.com/ISRCTN69073386.
110. Roh HC, Kim SJ, Kang SW, et al. Long-term outcomes of
polypoidal choroidal vasculopathy in comparison with typical exudative
age-related macular degeneration. Graefes Arch Clin Exp Ophthalmol. 2022;260:83–92.
111. Cheung CMG, Tan CS, Patalauskaite R, et al. Ranibizumab with or without verteporfin photodynamic therapy for
polypoidal choroidal vasculopathy: predictors of visual and anatomical response in the EVEREST II study.
Retina (Philadelphia, Pa). 2021;41:387–392.
112. Tan CS, Lim LW, Ngo WK, et al. EVEREST report 5: clinical outcomes and treatment response of
polypoidal choroidal vasculopathy subtypes in a multicenter, randomized controlled trial. Invest Ophthalmol Vis Sci. 2018;59:889–896.
113. Tan CS, Ngo WK, Lim LW, et al. A novel classification of the vascular patterns of
polypoidal choroidal vasculopathy and its relation to clinical outcomes. Br J Ophthalmol. 2014;98:1528–1533.
114. Kim H, Lee SC, Kwon KY, et al. Subfoveal choroidal thickness as a predictor of treatment response to anti-vascular endothelial growth factor therapy for
polypoidal choroidal vasculopathy. Graefes Arch Clin Exp Ophthalmol. 2016;254:1497–1503.
115. Jiménez-Santos M, Saenz-Francés F, Calvo-González C, et al. Subfoveal choroidal thickness as a potential predictor of treatment response after intravitreal ranibizumab injections for
polypoidal choroidal vasculopathy. Can J Ophthalmol. 2022. [Epub ahead of print].
116. Chang YC, Cheng CK. Difference between pachychoroid and nonpachychoroid
polypoidal choroidal vasculopathy and their response to anti-vascular endothelial growth factor therapy.
Retina (Philadelphia, Pa). 2020;40:1403–1411.
117. Kim YH, Lee B, Kang E, et al. Choroidal thickness profile and clinical outcomes in eyes with
polypoidal choroidal vasculopathy. Graefes Arch Clin Exp Ophthalmol. 2021;259:1711–1721.
118. Yanagi Y, Ting DSW, Ng WY, et al. Choroidal vascular hyperpermeability as a predictor of treatment response for
polypoidal choroidal vasculopathy.
Retina (Philadelphia, Pa). 2018;38:1509–1517.
119. Kim JY, Son WY, Kim RY, et al. Recurrence and visual prognostic factors of
polypoidal choroidal vasculopathy: 5-year results. Sci Rep. 2021;11:21572.
120. Cho HJ, Kim K, Lim SH, et al. Retinal pigment epithelial atrophy after anti-vascular endothelial growth factor therapy for
polypoidal choroidal vasculopathy. Br J Ophthalmol. 2020;104:1443–1447.
121. Choi EY, Park SE, Lee SC, et al. Long-term incidence and growth of chorioretinal atrophy in patients with
polypoidal choroidal vasculopathy. Ophthalmologica. 2020;243:136–144.
122. Mori R, Tanaka K, Yuzawa M. Factors predicting 2-year treatment results of ranibizumab therapy for
polypoidal choroidal vasculopathy in eyes with good baseline visual acuity. Medicine. 2018;97:e11188.
123. Asano-Shimizu K, Asano S, Murata H, et al. Early changes of vascular lesions and responses to combined photodynamic therapy in patients with
polypoidal choroidal vasculopathy. Int Ophthalmol. 2020;40:1335–1345.
124. Díaz-Villamarín X, Blánquez-Martínez D, Pozo-Agundo A, et al. Genetic variants affecting anti-VEGF drug response in
polypoidal choroidal vasculopathy patients: a systematic review and meta-analysis. Genes. 2020;11:1335
125. Yoneyama S, Sakurada Y, Kikushima W, et al. Genetic factors associated with response to as-needed aflibercept therapy for typical neovascular
age-related macular degeneration and
polypoidal choroidal vasculopathy. Sci Rep. 2020;10:7188.
