Abstract
Purpose
Polypoidal choroidal vasculopathy (PCV) may have frequent recurrences after fluid resolution, but time to recurrence is unclear. This study explored time to first polypoidal recurrence after 1-year fixed-dosing aflibercept treatments.
Design
Retrospective cohort study.
Methods
Treatment-naïve PCV eyes treated between April 2015 to May 2019 were identified and included with criteria including: (1) received fixed-dosing 2 mg aflibercept in the first year, (2) became “inactive” (absence of both intraretinal and subretinal fluid on OCT) at post-treatment year-1 (PTY1) and managed as needed (PRN) thereafter, (3) FU ≥ 12 months after PTY1. Fundus photography, indocyanine green angiography (ICGA), and OCT graded to identify timing and risk factors for recurrence (defined as fluid on OCT).
Results
Of 37 study eyes [37 patients; median age was 64 years (IQR 59–69); median aflibercept injection number was 8 (IQR 8–8); median FU 38 months (IQR, 30–50 months)]; 18 eyes (49 %) had recurrence during FU. Fourteen (78 %) of 18 had recurrence within 12 months after PTY1 visit. Risk factors for recurrence included: incomplete polypoidal regression on post-treatment ICGA [ P = .004, Hazard ratio (HR) = 4.4, 95 % confidence interval (CI) 1.6–11.9] and PED with internal heterogeneous reflectivity on post-treatment OCT ( P = .04, HR = 2.7, 95 % CI 1.1–6.9).
Conclusions
Nearly half of inactive PCV eyes following 1-year fixed-dosing aflibercept treatments had recurrent polypoidal lesions. Eyes with high-risk features for recurrence, some of which can be detected with OCT without the need for ICGA, may warrant close monitoring.
1
Introduction
Polypoidal choroidal vasculopathy (PCV) is a subtype of macular neovascularization type 1, which is reported more commonly among patients of Asian or African descent compared with White patients. PCV lesions comprise an abnormal branching choroidal vascular network and polypoidal lesions at the edge of the vascular network. Patients with PCV often present with central scotomas due to submacular hemorrhage presumably associated with rupture of polypoidal lesions, or decreased vision due to subretinal fluid associated with exudative leakage of polypoidal lesions or branching vascular network (BVN), which extend to the subfoveal region. Without timely management, this condition could lead to permanent visual loss.
Recommended first-line treatments for PCV include intravitreal anti-vascular endothelial growth factor (anti-VEGF) monotherapy or combined intravitreal anti-VEGF therapy and photodynamic therapy (PDT). Both therapeutic options have been proven beneficial in large randomized clinical trials with, on average, substantially improved visual acuity and anatomical outcomes. Apart from the Randomized clinical trials (RCTs), observational studies in the clinical practice setting, so-called “real world” studies, also have demonstrated that PCV eyes could achieve meaningful visual and anatomical outcomes following anti-VEGF injections with or without PDT. However, most RCTs and clinical practice setting studies follow each participant for only one year or two years after treatment initiation. If a participant leaves the study and treatments are discontinued, a recurrence of PCV might occur as PCV is known to have a high rate of recurrence after treatment has been withheld because the eye was stable functionally and anatomically.
Recurrence rates of PCV have been reported between 64 % and 79 % in PCV eyes receiving PDT monotherapy, and between 48 % and 82 % in PCV eyes receiving anti-VEGF monotherapy. While physicians are aware of the high recurrence of PCV, information regarding the timing of recurrence and risk factors for recurrence are unclear. Thus, this study aimed to identify time to first recurrence of inactive PCV eyes that had completed 1-year fixed-dosing aflibercept treatments and to determine risk factors for recurrence of PCV.
2
Methods
This study was a retrospective chart and imaging review. The study design had been approved by the Institutional Review Board and Ethical Committee of Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand, and followed the principles of the Declaration of Helsinki. Adult patients who were diagnosed with PCV, received treatments and follow-up (FU) care at Chiang Mai University Hospital between April 1, 2015 and August 31, 2022 were identified from medical records.
Inclusion criteria included the following: (1) treatment-naïve PCV eyes that completed 1-year fixed-dosing 2-mg aflibercept treatment — 3 monthly injections, then every 8 weeks; or as often as monthly if intraretinal (IRF) or subretinal (SRF) fluid persisted on optical coherence tomography (OCT); note that rescue photodynamic therapy (PDT) could be given per physicians’ discretion when IRF or SRF persisted despite 6 consecutive monthly aflibercept injections; (2) a treated eye became “inactive” (absence of both IRF and SRF on OCT) at post-treatment year-1 (PTY1) visit and managed as PRN regimen thereafter (FU with no treatment unless IRF or SRF or new subretinal hemorrhage was seen on OCT); (3) had FU duration of ≥ 12 months after PTY1 visit; and (4) had available indocyanine green angiography (ICGA) of the baseline and PTY1 visits, and OCT of all FU visits. Color fundus photography was also required at baseline but not mandatory at PTY1 and FU visits. If both eyes of the same patient were eligible, only the right eye would be the study eye. Exclusion criteria included the following: (1) presence of co-existing retinal conditions, e.g., diabetic retinopathy or retinal vascular occlusion; (2) presence of poor-quality imaging due to ocular media or other causes.
De-identified images of study eyes were sent to 2 teams of standardized image graders. Grader Team A (2 retinal specialists: VC, DP) reviewed pre-treatment colour fundus photography (CFP) and ICGA to confirm PCV diagnosis based on the criteria proposed by the efficacy and safety of verteporfin photodynamic therapy in combination with ranibizumab or alone versus ranibizumab monotherapy in patients with symptomatic macular polypoidal choroidal vasculopathy (EVEREST) study. The diagnosis was made when a hyperfluorescent nodule was seen on ICGA within the first 6 minutes, along with ≥ 1 of the following findings: subretinal orange nodule on fundus examination corresponding to the nodule on ICGA, massive subretinal hemorrhage, hypocyanescent halo around the nodule, nodular appearance on stereoscopic viewing, pulsatile filling of the nodule, or abnormal vasculature supplying the nodule. Also, grader team A reviewed PTY1 ICGA to determine if the study eye had complete or incomplete polypoidal regression by comparing the post-treatment images side-by-side to the pre-treatment ICGA images and also identified if BVN still presented on the post-treatment images. Complete polypoidal regression was defined as a complete disappearance of hyperfluorescent lesions on ICGA previously identified at baseline. Incomplete polypoidal regression was defined as an incomplete disappearance or persistent hyperfluorescent lesions on ICGA previously identified at baseline.
Grader team B (2 retinal fellows: TS, CN) confirmed PTY1 OCT images of all study eyes that all study eyes were “inactive” (no IRF or SRF on all b-scan OCT images), identified other OCT characteristics at post-treatment year-1 visit e.g., presence of double-layer sign, presence of pigment epithelial detachment (PED), internal reflectivity of PED, or presence of pachychoroid. Double-layer sign was defined as a thin separation between retinal pigment epithelium (RPE) and Bruch’s membrane. PED was defined as a dome-shaped RPE elevation with a height of ≥ 100 μm. The Internal reflectivity of PED was classified as homogeneous hyperreflectivity, homogeneous hyporeflectivity, and heterogeneous reflectivity ( Fig. 1 ). If an eye had more than one PED, homogeneous hyperreflectivity or hyporeflectivity only was used if all PEDs in the eye were homogeneously hyperreflective or hyporeflective. Pachychoroid was defined as a presence of thick choroid (subfoveal choroidal thickness exceeded the thickness of normal retina of the same eye) along with dilated haller layer vessels and attenuated overlying choriocapillaris. Grader team B also identified if there was any “recurrence” during FU by reviewing OCT and fundus photography of all FU visits of each study eye. “Recurrence” was defined as a presence of IRF or SRF on any b-scan OCT image, or a presence of new subretinal hemorrhage on fundus photography during follow-up visits. “Time to recurrence” was defined as an interval between “PTY1 visit” and “the first FU visit in which the recurrence was identified”.
For OCT images, 49 b-scan/macular cube (Spectralis®, Heidelberg Engineering, Heidelberg, Germany) scans were performed routinely in the clinic and evaluated in this study. Disagreements between graders in each team were managed by open adjudications.
Demographic characteristics were analyzed using descriptive statistics, presenting median and interquartile range (IQR) for continuous data, and frequency with percentage for categorical data. Visual acuity (VA) was measured in Early Treatment Diabetic Retinopathy Study (ETDRS) letter scores. Additionally, Kaplan-Meier survival analysis was applied to determine the time to recurrence. To adjust for differing follow-up durations among groups, survival differences were analyzed using a Cox proportional hazards regression model.
To evaluate univariable associations and identify potential risk factors for recurrence, a generalized linear model (GLM) with a binomial family was employed. The factors subjected to analysis included characteristics observed in OCT and ICGA at PTY1 visit. Post-treatment ICGA-related features included the status of polypoidal regression, presence of BVN, and total PCV lesion size. Post-treatment OCT characteristics comprised the presence of PED, PED characteristics, total number of PED occurrences, presence of BVN, total PCV lesion size, and presence of pachychoroid. All statistical analyses were conducted using Stata® (version 17.0; StataCorp), and all P values were two-sided but were not adjusted for multiple analyses.
3
Results
There were 58 treatment-naïve PCV eyes receiving fixed-dosing aflibercept treatments with rescue verteporfin photodynamic therapy (vPDT) during the study period. Twenty-one eyes were excluded as post-treatment ICGA images were not available (10 eyes), FU duration of less than 12-month (7 eyes), missing OCT images at some FU visits (1 eye), and presence of co-existing retinal diseases i.e., vitreoretinal interface abnormalities (2 eyes), branch retinal artery occlusion (1 eye).
Therefore, this study included a total of 37 study eyes (37 patients; 56 % were female, median age was 64 years — interquartile range (IQR), 59–69 years. Median presenting visual acuity ETDRS letter score at baseline was 45 (IQR, 25–55) (Snellen equivalent 20/125 IQR, 20/63–20/320). Total FU ranged from 21 to 72 months; median FU duration was 38 months (IQR, 30–50 months). Median duration between each FU visit was 2 months (IQR, 1.5–4 months). All study eyes received intravitreal aflibercept treatments, the injection number ranged from 8 to 12 injections during the 1-year treatment period (median injection number was 8 injections; IQR, 8–8 injections), and only 2 eyes (5 %) required rescued PDT. The OCT findings at PTY1 visit showed an absence of both IRF and SRF in all study eyes (37 of 37, 100 %). The Kaplan-Meier failure estimate illustrated that the probability of recurrence approached 50 % by 18 months after baseline ( Fig. 2 ).
3.1
PCV eyes with recurrences
Of the 37 study eyes, 18 (49 %) had recurrence during FU. Of these 18 eyes, 10 (55 %) had recurrence during the first 6 months (month 0–6 from PTY1), 4 (22 %) had recurrence during month 7–12, 3 (16 %) had recurrence during month 13–24, and 1 eye (5 %) had recurrence during month 25–36 of FU ( Fig. 3 ). Therefore, 14 (78 %) of 18 eyes had recurrence during the first year after treatment deferral. Median follow-up duration was 43.5 months (IQR, 34.5–53 months). Duration between each follow-up visit ranged from 1 to 6 months and the median was 2 months (IQR, 2–4 months).
At the PTY1 visit, among 18 eyes with recurrence, 12 (66 %) had incomplete polypoidal regression and 6 (33 %) had complete polypoidal regression on the post-treatment ICGA. Examples of PCV eyes with incomplete and complete polypoidal regression at PTY1 visit that developed recurrence of PCV were demonstrated in Fig. 4 and Fig. 5 , respectively.
