Polypoidal choroidal vasculopathy (PCV) is characterised by a complex of branching vascular networks terminating in aneurysmal or polypoidal lesions that appear as reddish orange structures.1^–3 PCV often accompanies fluid or haemorrhage at the subretinal or subretinal pigment epithelial region and affects vision.4 Previous studies have reported a higher prevalence of PCV in patients presumed to have age-related macular degeneration (AMD) in Asian populations than in Caucasians.4^–6

We previously reported that photodynamic therapy (PDT) for PCV was more effective in maintaining or improving vision than for choroidal neovascularisation (CNV) resulting from AMD 1 year after treatment.5 However, we also showed that recurrent or newly developed polypoidal lesions affect vision over several years’ follow-up after PDT for PCV.7 To determine the role of PDT in treating the vascular abnormalities of PCV, the changes in the vascular component through the course after PDT should be evaluated.

Indocyanine green angiography (ICGA) taken by confocal scanning laser ophthalmoscopy (SLO) provides higher-resolution images than a fundus camera and allows visualisation of even faint vascular changes in the PCV lesion.8 Optical coherence tomography (OCT) shows the vascular abnormalities as characteristic protrusions of the retinal pigment epithelium (RPE) at the polypoidal lesion9 10 and double reflective layers at the area with branching vascular network as well as the accompanying fluid.11 Considering the findings obtained by these technologies together, the changes in the vascular component of PCV can be monitored precisely.

The aim of the present study was to determine the angiographic and tomographic changes of PCV over time in a consecutive series of patients with PCV treated with PDT.

PATIENTS AND METHODS

The study was a prospective, consecutive, interventional case series conducted at Osaka University Hospital after the approval of the institutional review board was obtained. Clinical records and angiographic findings of 31 consecutive eyes of 31 patients with PCV involving the fovea and treated with PDT between July 2005 and August 2006 were evaluated. Some cases had been included in a previous study.8 All patients completed at least 15 months of follow-up.

The diagnosis of PCV was established based on the orange-red vascular lesion on fundus examination and the finding of polyp-like choroidal vessel dilatation with or without a branching vascular network on ICGA using SLO (Heidelberg Retina Angiograph 2 (HRA2), Heidelberg Engineering, Heidelberg, Germany). The findings from OCT also were used for the diagnosis of PCV such as protruding retinal pigment epithelium (RPE) and a double layer sign corresponding to the polypoidal lesion and branching vascular network, respectively.

All patients underwent a comprehensive ocular examination, including fundus examinations, colour fundus photography, OCT (Stratus OCT, Carl Zeiss Meditec, Dublin, CA), fluorescein angiography (FA), and ICGA at baseline and every 3-month visit within 24 months. If active lesions were observed at 24 months, those examinations were repeated every 3 months. ICGA was performed using confocal SLO. OCT was performed using the line scan mode. PDT with verteporfin (Visudyne, Novartis) was administered as an ICGA-guided method as previously reported;5 therefore, an adjacent RPE detachment or haemorrhage was not included within the lesion. We applied PDT separately when the eyes had more than one polypoidal lesion, and the PDT spot was expected to include an area of normal choroidal vasculature that was larger than that of the PCV lesions as a whole. Each exposure was applied for 83 s successively and not overlapped. The purpose of this approach was to minimise the possibility of choroidal ischaemic changes. Additional treatments were performed under ICGA guidance when the exudation from the recurrence of a polypoidal lesion or vascular abnormalities affected the vision. The recurrence of polypoidal lesions was defined as the reappearance of active polypoidal lesions observed on ICGA and FA after resolution of the initial polypoidal lesions.

The major outcome measures were changes in the findings on ICGA and OCT during the follow-up period after PDT. The resolution and recurrence of polypoidal lesions or specific vascular changes on ICGA were noted. Associated OCT features such as RPE protrusion or the presence of fluid also were evaluated. The planimetric size of the branching vascular network was measured from early phase ICGA at baseline and the final visits using the software included in the HRA2. A decrease or increase in the lesion size was recognised when the size changed by more than 20% at the corresponding area. The visual outcomes also were recorded. Statistical evaluation was performed with SigmaStat software version 3.1 (SPSS, Chicago).

RESULTS

Thirty-one eyes of 31 patients (23 men, eight women) were included; the mean follow-up was 19.2 (SD 4.8) months (range 15–30). The mean age was 70.8 (7.7) years (range 53–83). The mean number of treatments per eye during the follow-up period was 2.1 (0.9) (range 1–4). The best-corrected visual acuity (BCVA) before treatment ranged from 0.1 to 0.7 (mean 0.34) and improved to 0.45 (p = 0.003) and 0.47 (p = 0.013) at 12 months and the final visit, respectively. Twelve eyes (39%) had an improvement in the BCVA of 0.3 logMAR score or more; three eyes (10%) had a decrease in the BCVA by 0.3 logMAR score or more at the final visit.

At baseline, polypoidal lesions with a complex of branching vascular networks were detected by ICGA in all eyes. Steeply protruding RPE corresponding to the polypoidal lesions was confirmed on OCT in all eyes. OCT showed the presence of subretinal fluid (SRF) in 30 eyes (97%), a serous and/or haemorrhagic pigment epithelial detachment (PED) in 12 eyes (39%), and cystic macular oedema in two eyes (6%). Fibrin deposits were suspected in eight eyes (26%). Eight (26%) eyes had multiple polypoidal lesions, whereas 23 (74%) eyes had one lesion. PDT was performed separately in four eyes with separate polypoidal lesions.

After PDT, polypoidal lesions were not apparent on ICGA at at least one visit in 29 eyes (94%) with the absorption of the accompanying fluid. The protruding RPE corresponding to the polypoidal lesion decreased in height or flattened when the polypoidal lesions regressed. The branching vascular network remained in all 31 eyes throughout follow-up. The size of the branching vascular network increased in 13 eyes (42%), remained unchanged in 14 eyes (45%) and decreased in four eyes (13%) at the final visit compared with baseline. In the eyes in which the branching vascular network increased in size, the vessels extended beyond the original polypoidal lesions (figs 1, 2) without exudation.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Findings of indocyanine green angiography (ICGA) and optical coherence tomography (OCT) at the macula in a 60-year-old man (A–D) and a 71-year-old man (E–H) before and after PDT. (A) Baseline ICGA shows two regions consisting of multiple polyps (arrows) and branching vascular networks (arrowheads). The horizontal OCT shows a large protrusion of the retinal pigment epithelium (RPE) corresponding to the multiple polyps at the macula and subretinal fluid (SRF). (B) Three months after the first PDT treatment, one polypoidal lesion has completely regressed, and OCT shows a flattened RPE and resolution of SRF. Another PDT treatment was administered to the remaining polypoidal lesion. (C) Twelve months after the first PDT, both polypoidal lesions regressed and the branching network extended toward the area once occupied by the polypoidal lesions. (D) Twenty-four months after PDT, branching vascular networks (arrowheads) without exudation continued to increase in size. The VA improved from 0.1 at baseline to 0.4 at 24 months. (E) ICGA before PDT shows polypoidal lesions (arrow) and branching vascular networks (arrowheads). The vertical OCT confirms the pigment epithelial detachment (PED) and SRF. The VA is 0.5. (F) Three months after PDT, the polypoidal lesion remains on ICGA and PDT was performed again. OCT still shows PED and SRF. (G) Six months after PDT, the PED has been replaced by the branching vascular networks. OCT shows resolution of PED and SRF. (H) Twenty-four months after PDT, the branching vascular networks increased in size (arrowheads). Presumed recurrence of the polypoidal lesion inferior to the fovea is observed at the termini of branching vascular networks (arrow) without exudation. OCT shows slight elevation of the RPE. The VA remains 0.5. Patient consent has been obtained for publication of this figure.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Findings in a 65-year-old man. (A) Baseline indocyanine green angiography (ICGA) and a vertical optical coherence tomography (OCT) image show branching vascular networks (arrowheads) terminating in polypoidal lesions (arrows). OCT shows protruded RPE corresponding to the polypoidal lesions accompanied by a serous pigment epithelial detachment (PED). Subretinal fluid (SRF) is also seen. The VA is 0.5. The patient underwent PDT at this visit and 3 months later. (B) Six months after PDT, ICGA shows gradual regression of polypoidal lesions at the initial site. OCT shows a decrease in the height of the PED. Another PDT was performed at this visit. (C) Examination 9 months after the initial treatment shows complete regression of the polypoidal lesions; the VA improved to 1.0. The PED and SRF have resolved completely on OCT. (D) The polypoidal lesions (arrows) recurred 12 months after the initial PDT at the termini of the extended branching vascular networks (arrowheads). Recurrent SRF is seen on OCT. The VA deteriorated to 0.7. Patient consent has been obtained for publication of this figure.

During follow-up, polypoidal lesions reappeared in 10 of 29 eyes (34%), at the initial site in five eyes, at a different site connected to the branching vascular networks in four eyes (fig 2) and in both regions in one eye. The mean period between the initial PDT and the first detection of the recurrent or the newly developed polypoidal lesions was 11.4 (1.9) months (range 9–15 months). The baseline age, VA, lesion size and number of polypoidal lesions did not differ significantly between the group in which the lesions recurred and the group in which they did not recur (p = 0.908, p = 0.472, p = 0.183, p = 0.718, respectively). When recurrent or newly developed polypoidal lesions were detected, the SRF recurred in eight eyes (80%), the PED increased or recurred in six eyes (60%), and new SRH was observed in five eyes (50%). All 10 eyes with reappearance of active polypoidal lesions underwent additional PDT. After PDT, the polypoidal lesions resolved again in eight eyes with the resolution of the fluid and remained in two eyes. The mean VA did not differ significantly (p = 0.298) at the final visit in eyes in which the polypoidal lesion did and did not recur.

Abnormal choroidal vessels, which were distinguishable from the characteristic branching vascular networks and not a polypoidal vasculature, became apparent on both ICGA and OCT after PDT at the initial polypoidal lesions in seven (23%) eyes, that is, thin curled choroidal vessels with fibrinous exudation in four eyes (fig 3A–D) and radiating choroidal vessels in three eyes (fig 3E–L). Thin curled choroidal vessels were seen at 6 months after the initial application of PDT in all four eyes. OCT showed SRF, intraretinal cysts and a fluctuating RPE layer with subretinal material, and FA showed classic CNV-like leakage from the regions corresponding to the abnormal vessels. Another three eyes with radiating choroidal vessels showed expanding vascular networks with fewer branches with apparent elevations of the RPE similar to a vascularised PED, with an orange-red appearance at the initial polypoidal lesions at 9 months (one eye) and 12 months (two eyes) after the first treatment. FA showed occult CNV-like leakage at the regions of PED. Additional treatment was applied in four eyes with visual deterioration, and the reduction or resolution of vascular abnormalities was seen in two eyes, although there was a recurrence in one of these eyes (fig 3A–D). In the other two eyes, additional treatment did not affect the vascular abnormalities, although the fluid decreased.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3 Findings on indocyanine green angiography (ICGA) and optical coherence tomography (OCT) before and after photodynamic therapy (PDT) in a 72-year-old man (A–D), another 72-year-old man (E–H) and an 80-year-old man (I–L). (A) ICGA before PDT shows small polypoidal lesions and branching vascular networks. The vertical OCT at the fovea shows protruding RPE corresponding to the polypoidal lesion. The VA is 0.7. (B) Findings 6 months after PDT. Although the polypoidal lesion is not apparent on ICGA, thin curled choroidal vessels at the subfoveal area are visible (arrowheads). OCT shows fibrin or presumed neovascular components in the subretinal space. Additional PDT was applied. (C) Nine months after the initial application of PDT, the abnormal choroidal vessels have decreased in size. The subretinal materials or possible neovascular components have regressed with the elevated RPE on OCT. (D) Twelve months after PDT, the size of the neovascular networks (arrowheads) on ICGA has increased. OCT shows the recurrence of subretinal materials or presumed neovascular components and SRF. The VA decreased to 0.4. (E) ICGA shows polypoidal lesions and small branching vascular networks at the fovea; the horizontal OCT image confirms the protruding retinal pigment epithelium (RPE), accompanying pigment epithelial detachment (PED) and subretinal fluid (SRF). After the PDT, subpigment epithelial haemorrhage had occurred. (F) Three months after PDT, the polypoidal lesion resolved on ICGA; however, OCT shows residual RPE protrusion. (G) Twelve months after PDT, the abnormal choroidal vasculature progressed toward the area where the polypoidal lesions had existed (arrowheads) and OCT shows the PED extending laterally. (H) Eighteen months after PDT, the abnormal choroidal vasculature has continued to increase in size (arrowheads). OCT shows increased PED height and extent without visual deterioration. (I) ICGA before PDT shows two polypoidal lesions and branching vascular networks. OCT confirms the protrusion of the RPE and SRF. (J) Three months after PDT, the polypoidal lesions disappeared on ICGA. Newly irregular networks are evident (arrowheads) and PDT was performed again during this visit. OCT shows PED and SRF. (K) Six months after PDT, the PED has been replaced by the vascular networks. OCT shows resolution of the PED and SRF. (L) Twelve months after PDT, the polypoidal lesion temporal to the fovea has recurred. Thin radiating vessels connected to the branching vascular networks extend involving the macula (arrowheads). OCT shows slight elevation of the RPE with subretinal exudation (arrow). Patient consent has been obtained for publication of this figure.

DISCUSSION

Results from recent clinical trials have shown a beneficial effect of PDT on VA for symptomatic PCV.5 12^–15 We have shown that PDT is significantly more efficacious for PCV than for AMD, resulting in improved VA and resolution of exudative changes in most cases at 1 year; however, ICGA at 1 year showed recurrent polyps in 9% of eyes at the same or different site.5 The possibility that a longer follow-up period might link the higher recurrence rate of polypoidal lesions has been postulated;7 13 therefore, it should be clarified how the PCV vascular lesion changes after PDT sequentially.

Thus, in the present study, we focused on investigating angiographic changes after PDT using confocal SLO and the associated findings on OCT. Because our previous study showed that the confocal SLO is superior for detecting the thin vessels of vascular networks of the PCV lesion than a fundus camera,8 we performed this study using confocal SLO. Since the OCT findings for the vascular lesions of polypoidal dilatation and branching vascular network are well known, the changes on OCT also were evaluated.9^–11

Polypoidal lesions initially resolved at a rate of 94% after PDT, which was similar to the recent reports in Japan.5 14 15 The protruding RPE on OCT corresponding to the polypoidal lesions flattened when complete resolution was observed on ICGA. During the study periods, recurrent or newly developed polypoidal lesions were observed in 10 (32%) eyes. The high recurrence rate might be due to the enhanced visualisation of the changes in the choroidal vasculature using confocal SLO. When the polypoidal lesions recurred, SRF or a PED was detected on OCT in most cases; therefore, careful evaluation of OCT findings may help detect the recurrence.

All the recurrent polypoidal lesions were connected to the branching vascular networks. Yuzawa et al16 reported that residual branching vascular networks after laser photocoagulation were associated with disease recurrence. The branching vascular networks remain even after laser exposure of PDT, as Akaza et al recently reported.15 Therefore, if PDT cannot resolve the branching vascular networks of PCV, recurrence might be inevitable. Unexpectedly, the region of the branching vascular networks progressed in 13 (42%) eyes after PDT. The reason for this is unclear; however, the vessels seemed to extend toward the absorbed PED or resolved polypoidal lesions, so morphological abnormalities between the RPE and inner choroidal vasculature may induce this. It is also possible that ischaemic changes in the surrounding choroidal region included by the PDT application induce enlargement of the branching vascular network. Further studies are needed to clarify this.

By detailed examination of the findings on ICG, we observed an abnormal choroidal vasculature that differed from the characteristic vascular component of PCV at the initial polypoidal lesions in seven eyes (23%), that is, thin curled vessels in four eyes and radiated vessels with elevated RPE in three eyes. Thin curled vessels showed classic CNV-type leakage on FA and accompanied SRF, intraretinal cysts, and subretinal fibrin confirmed on OCT (fig 3A–D). Those findings seemed similar to that observed in type II CNV.17 18 The radiating vessels had irregularly elevated RPE on OCT similar to findings associated with occult (type I) CNV (fig 3E–L).19 From these findings, it was suggested that the vascular lesions of PCV might evolve to CNV after PDT.

Previous studies have reported that concomitant type II CNV in eyes with PCV is present5 20 and some eyes with PCV with fibrin deposits are difficult to differentiate from type II CNV;20 therefore, changes in those vascular findings may be due to the appearance of masked CNV after resolution of the PCV. Some of these abnormal vessels may be immature recurrent polypoidal lesions. However, in addition to the current study in which the presumed CNV appeared after PDT for PCV, our previous study had cases with subsequent development of PCV after PDT for AMD.5 Therefore, we consider the possibility that PCV and AMD in Japanese patients may share common characteristics. The LOC387715/HTRA1 variants recently have been shown to be associated with both PCV and AMD in a Japanese population.21 Thus, PCV and AMD share common genetic factors, which may support our speculation that PCV and AMD share some similar clinical features.

As the result of the disappointing effects of anti-VEGF therapy for PCV,22 PDT might be the first treatment choice for PCV based on its current efficacy, but PDT does not seem to stabilise the disease permanently. The changes in the remaining branching vascular network should be monitored closely, even after the polypoidal lesions resolve.