1.Introduction

---

Idiopathic choroidal neovascularization (CNV) is defined as CNV in patients younger than 50 years, without any apparent primary ocular or systemic diseases.[1,2] The natural history and final visual outcomes of idiopathic CNV are generally considered to be more favorable than CNV due to age-related macular degeneration (AMD) [3,4]. However, the natural course of the disorder can be unpredictable, and treatment is recommended for these patients [1,3,4]

Various treatment modalities have been used for idiopathic CNV. Because idiopathic CNV usually appears as classic CNV, photodynamic therapy (PDT) was recommended as mainstay of treatment for idiopathic CNV before the introduction of anti-vascular endothelial growth factor (VEGF) agents [5, 6]. Even in the era of anti-VEGF, PDT is still considered a good treatment option for some diseases, including polypoidal choroidal vasculopathy (PCV) [7, 8] A recent retrospective study demonstrated favorable visual outcomes after PDT for PCV with long-term follow-up periods over 60 months [9]. Long-term follow-up study for myopic CNV also showed a stabilizing effect of PDT, suggesting PDT as an attractive option, especially for juxtafoveal myopic CNV [10].

Long-term treatment efficacy of PDT has not been investigated for idiopathic CNV, although it was considered as a good treatment option for idoipathic CNV. In this study, we investigated the 5-year visual outcomes of eyes with idiopathic CNV after PDT, to determine the long-term efficacy and safety of PDT. In addition, additional analysis was performed to find out visual prognostic factors after PDT in these patients.

2.Methods

---

The medical records for 30 eyes of 27 patients with naive idiopathic CNV were retrospectively reviewed. The first treatment of each patient was performed between March 2005 and March 2007, and all included patients completed at least 5 years of follow-up at Yonsei University Medical Center. Informed consent was obtained from all eligible participants at the time of treatment, and this study was approved by the Institutional Review Board of Yonsei University Medical Center.

Inclusion criteria for idiopathic CNV were: (1) age <50 years; and (2) absence of concurrent ocular diseases in the study eyes that compromised or could have compromised vision, such as pathologic myopia.. Exclusion criteria were: (1) history of prior treatment for CNV (including laser, submacular surgery, PDT, or intravitreal anti-VEGF therapy); (2) extrafoveal location of CNV.

The patients’ characteristics were retrieved from the medical charts, including age at initial diagnosis, gender, refractive errors, greatest linear diameter at baseline, and best-corrected visual acuity (BCVA) determined using Decimal charts.. Decimal BCVA results were converted to a logarithm of the minimum angle of resolution (logMAR) value for statistical analysis. The refractive errors for all patients were measured by an autorefractometer without cycloplegia. We further divided the patients into two groups according to the location of CNV lesion: juxtafovea and subfovea, and compared visual outcome between two groups.Greatest linear dimension at baseline was measured manually at the early phase of ICGA, using software embeeded in the Heidelberg Retinal Angiograph system (HRA-2; Heidelberg engineering, Dossenheim, Germany).

PDT with verteporfin was performed according to previous studies.[11,12] We defined the recurrence of idiopathic CNV as the reappearance of active CNV lesions with leakage on FA after more than 6 months without treatment. Retreatment for recurrence was considered based upon FA, ICGA, and optical coherence tomography findings.

The patients were observed 1 month after PDT, and at 1- to 2-month intervals based on the patients' condition during the first year of treatment. When there was no evidence of active disease, annual follow-up was done for the patients. At every visit, BCVA testing, slit lamp examination, and dilated fundus examination were performed. In cases of changing symptoms, including vision worsening or metamorphopsia, additional FA, ICGA, and optical coherence tomography were performed to rule out recurrence of CNV.

The primary measured outcomes were the mean BCVA from baseline at each follow-up. We also defined visual gain as 0.3 logMAR or more visual improvement, and visual loss as visual deterioration of 0.3 logMAR or more when compared with baseline.

The Kolmogorov-Smirnove test was used for verification of normal distribution in this study group. Paired t-test was used to compare the baseline BCVA to that at each visit. In addition, non-parametric analyses including the Mann-Whitney U test and Wilcoxon signed-rank test were used for subgroup analyses. Stepwise multiple regression analysis was performed to find out predictive factors for visual outcome. Statistical analysis was performed using SPSS 18.0 software for Windows (SPSS Inc., Chicago, IL, USA). P values of less than 0.05 were considered statistically significant.

3.Results

3.1.Characteristics of the Patients

---

Mean age at diagnosis was 36.73±6.15 years, and 19 patients (63.3%) were female. All eyes showed type 2 CNV by FA, and distinct dark rims surrounding CNV lesions by ICGA. The greatest linear diameter was 947.16±484.95 µm at baseline. During mean follow-up of 75.97±15.50 months, a total mean number of PDT was 1.87±1.31 times.. 8 eyes (26.7%) experienced at least one recurrence, with mean number of recurrence as 1.33±0.71 times. The mean interval between first remission and first recurrence was 21.57±2.37 months. Mean number of PDT treatments was 2.89±1.62 times in the eyes with recurrence. Patients’ characteristics were summarized in Table 1.

3.2. Visual Outcomes During Follow-Up Period after Photodynamic Therapy

---

Mean BCVA was 0.56±0.38 logMAR (20/72 Snellen equivalent) at baseline, and 0.56±0.86 logMAR (20/72 Snellen equivalent) at 5 years (P=0.980). Changes of mean BCVA during follow-up are depicted in Figure 1. There was no significant difference between mean BCVA at each follow-up and baseline (P=0.581 at 1 month, P=0.576 at 3 months, P=0.681 at 6 months, P=0.417 at 1 year, P=0.452 at 2 years, P=0.480 at 3 years, and P=0.822 at 4 years).

BCVA at each follow-up were classified as visual gain and deterioration when compared with baseline BCVA (Figure 2). 14 eyes (46.7%) gained vision, whereas 6 eyes (20.0%) lost vision at 5-year follow-up visits. The 6 eyes with visual loss at 5 years showed disciform scarring at the fovea.

Figure1

Figure1. Changes of mean best-corrected visual acuity (BCVA; logarithm of the minimum angle of resolution, logMAR) from baseline and each follow-up in eyes with idiopathic choroidal neovascularization (CNV) after photodynamic therapy. There was no significant change at each follow-up visit when compared with baseline.

Figure2

Figure2. Number of eyes showing visual gain and visual loss after photodynamic therapy for idiopathic choroidal neovascularization. Visual gain was defined as improvement of best-corrected visual acuity (BCVA) 0.3 logarithm of the minimum angle resolution (logMAR) or more, and visual loss was defined as deterioration of BCVA 0.3 logMAR or more than baseline BCVA. Among overall patients, 14 eyes (46.7%) gained vision, whereas 6 eyes (20.0%) lost vision at 5 years.

3.3. Subfoveal Versus Juxtafoveal Idiopathic Choroidal Neovascularization

---

Mean BCVA at baseline was 0.67±0.39 logMAR (20/93 Snellen equivalent) in subfoveal group, and 0.44± 0.36 logMAR (20/55 Snellen equivalent) in juxtafoveal group (P=0.173). Mean BCVA at 5 years was 0.57± 0.91 logMAR in subfoveal group, and 0.55± 0.86 logMAR in juxtafoveal group (P=0.705). The changes of mean BCVA in each group were depicted in Figure 1.

The eyes of subfoveal and juxtafoveal groups were also classified into visual gain and visual deterioration according to the BCVA changes from baseline BCVA. At 5 years, 9 eyes (47.4%) of the subfoveal group and 5 eyes (45.4%) of the juxtafoveal group gained vision (Figure 3). 3 eyes of each group (subfoveal, 15.8%; and juxtafoveal, 27.3%) lost vision at 5 years (Figure 4). Representative case was shown in Figure 5.

Figure3

Figure3. Percentage of eyes with visual gain defined as visual improvement of 0.3 logarithm of the minimum angle resolution (logMAR) or more when compared with baseline after photodynamic therapy for idiopathic choroidal neovascularization (CNV). Among the eyes with subfoveal CNV, 47.4% (9 eyes) gained vision, and 45.4% (5 eyes) with juxtafoveal CNV gained vision at 5 years when compared with baseline.

Figure4

Figure4. Percentage of eyes with visual loss defined as visual deterioration 0.3 logarithm of the minimum angle resolution (logMAR) or more when compared with baseline after photodynamic therapy for idiopathic choroidal neovascularization (CNV). 15.8% (3 eyes) with subfoveal CNV, and 27.3% (3 eyes) with juxtafoveal CNV lost vision when compared with baseline.

Figure5

Figure5. A 40-year-old female patient visited the ophthalmology clinic, complaining decreased vision in the left eye for 1 month. Best-corrected visual acuity (BCVA) of the left eye was 0.4 logarithm of the minimum angle resolution (logMAR). The greatest linear diameter was 1.25 mm at baseline. Refractive error with spherical equivalent was -1.50 diopter in both eyes. (Top left) On fundus examination, well-demarcated subretinal gray membrane with exudative fluid was shown at juxtafovea. Otherwise, there was no remarkable findings in fundus of the left eye. (Top middle) Well demarcated juxtafoveal choroidal neovascularization was documented on the early phase fluorescein angiography (FA). (Top right) Late phase FA showed profuse leakage from neovascular membrane. (Bottm left) Well demarcated neovascular membrane was surrounded with dark rim on indocyanine green angiography. Conventional photodynamic therapy was performed involving the entire neovascular membrane. (Bottom right) 5 years after photodynamic therapy, fundus examination showed subretinal fibrosis involving fovea. BCVA of the left eye was count finger.

3.4. Ocular/Systemic Complications

---

None of the patients developed systemic or ocular complications associated with PDT or supplemental verteporfin.

4.Discussion

---

In this study, we evaluated the long-term efficacy of PDT for idiopathic CNV. At the end of follow-up, 46.7% gained vision, whereas 20.0% lost vision when compared with baseline. The results of current study indicated that PDT seemed to have a long-term stabilizing efficacy for idiopathic CNV. However, when compared with a natural history study which showed only 5% of eyes with significant visuasl loss,[1] long-term follow-up of PDT showed limited visual outcome. When classified the eyes into subfoveal and juxtafoveal, those with juxtafoveal CNV lost gain more than those with subfoveal CNV, despite of better baselilne BCVA. The results suggest that those with juxtafoveal CNV showed worse visual outcome when treated with PDT.

Limited long-term visual outcome, especially with those with juxtafoveal CNV, may be due to harmful effect of PDT on retinal pigment epithelium (RPE). Dark rim surrounding idiopathic CNV on ICGA corresponds to multilayered and proliferated RPE at the outer margin of the neovascular membrane [13-15] This proliferated RPE layer seems to be related to the recovery mechanism of RPE which suppresses neovascularization [16,17] However, PDT can compromise this recovery mechanism of RPE, leading to limited recovery in these patients. It has been well documented that standard PDT could induce various changes including RPE damage [18]. Various degrees of RPE damage and RPE atrophy following PDT also have been reported [19,20]. These deteriorating effects of PDT on RPE may lead to a limited long-term visual outcome in idiopathic CNV, especially in juxtafoveal CNV in this study.

We also investigated the prognostic factors for long-term visual outcome after PDT in idiopathic CNV. Results showed that baseline BCVA, initial lesion size (greatest linear diameter at baseline), and location of CNV were visual prognostic factors for visual outcome. It has been well documented that baseline lesion size and BCVA were significantly correlated with visual outcome in AMD patients,including PCV [21-23]. In addition, the results of our study showed that juxtafoveal location of CNV was associated with worse visual outcome than subfoveal CNV after PDT. This may be due to the harmful effect of PDT on foveal center in the patients with juxtafoveal CNV, with further compromising foveal function.

This study has several limitations, including its retrospective nature and the relatively small study population. In addition, anti-VEGF therapy has become the mainstay of treatment for idiopathic CNV [24,25]. When compared with PDT, anti-VEGF therapy showed better visual outcome [26]. However, we think validation of long-term visual outcome after PDT is also important in idiopathic CNV. Our results indirectly support the use of anti-VEGF therapy, and can be used as a control data when validating the long-term efficacy of anti-VEGF therapy.

In conclusion, PDT for idiopathic CNV showed limited long-term visual outcome. Baseline BCVA, initial lesion size, and location of CNV were significantly correlated visual long-term visual outcome in these patients.

4.Conclusion

---

RDW is a sensitive (79.46%) indicator in detecting or screening IDA thus aiding in early diagnosis of the children. So, RDW of more than 17.9% can be used as an effective tool for the diagnosis of IDA in a large number of samples at major hospitals thereby reducing the manpower and unnecessary time consumption.

5.Acknowledgement

---

a. Funding/Support: none
b. Financial Disclosure: none
c. Contributions of Authors: Design and conduct of study (HM Kang, HJ Koh); collection of data (HM Kang); Management, analysis, and interpretation of data (HM Kang, HJ Koh); and preparation (HM Kang, HJ Koh); review, and approval of the manuscript (HM Kang, HJ Koh)
d. Other Acknowledgments: none