Abstract
Purpose
This study aimed to assess the role of photodynamic therapy (PDT) as an adjunct to anti-vascular endothelial growth factor (Anti-VEGF) intravitreal injections in the treatment of neovascular age-related macular degeneration (nvAMD).
Methods
PubMed, Cochrane Library, and ClinicalTrials.gov were searched for keywords “macular degeneration” and “photodynamic therapy” and “placebo” or “ranibizumab” or “bevacizumab” or “aflibercept” from inception to 2023. Included studies were peer-reviewed primary data reporting 12-month treatment results of nvAMD with anti-VEGF and PDT, anti-VEGF alone, intravitreal triamcinolone, or placebo. 23 studies were included in the final analysis. The major outcomes were best-corrected visual acuity (BCVA), central retinal thickness (CRT), and injection burden at 12 months.
Results
Anti-VEGF + PDT had better BCVA at 12 months compared to anti-VEGF (MD −0.07; 95 % CI −0.12, −0.01; P = 0.02). There was no significant difference in CRT at 12 months in anti-VEGF + PDT group versus anti-VEGF (MD −3.66; 95 % CI −10.28, 2.98; P = 0.28). Anti-VEGF + PDT group had significantly fewer injections compared to anti-VEGF (MD −1.76; 95 % CI −1.95, −1.58; P < 0.0001). There was no significant difference in pooled ocular adverse events between anti-VEGF + PDT versus anti-VEGF (MD 0.96; 95 % CI 0.68, 1.36; P = 0.41).
Conclusions
PDT is a successful adjunctive to anti-VEGF injections for the treatment of nvAMD. The combination of the therapies leads to improved BCVA at 12 months, decreased injection burden, and no difference in ocular safety.
1
Introduction
Age-related macular degeneration (AMD) is the leading cause of blindness affecting 200 million individuals worldwide, with an expected increase to 300 million affected individuals by 2040 .
AMD is characterized by choroidal thinning, thickening of Bruch’s membrane, changes to the retinal pigment epithelium (RPE), and drusen formation, which can lead to progressive central vision distortion and loss. Advanced AMD occurs in two forms, exudative/neovascular (“wet”) and non-neovascular/atrophic (“dry”); both can cause central vision loss and consequently significant disruptions in quality of life . Neovascular AMD (nvAMD) is characterized by the presence of choroidal neovascularization (CNV) within the macula . The clinical manifestations of nvAMD include subretinal and intraretinal fluid, retinal hemorrhage, lipid exudates, and RPE detachment and tear, which can lead to the formation of the ‘disciform scar’ (hypertrophic, fibrovascular, or atrophic macular scar), causing central vision impairment .
Neovascular AMD is classified into subtypes depending on the location and pattern of neovascularization, as well as leakage of fluorescein and indocyanine angiography . Fluorescein angiography (FA) is used to define a lesion’s location and size, allowing for the distinction between well-demarcated classic CNV and ill-defined occult CNV . Classic CNV typically appears as central hypo-fluorescence in the early phase. It can also present with an atypical classic presentation, characterized by increased fibrosis and CNV with fluorescein leakage and staining in the early-mid phase . Occult CNV is variable in presentation but can be further characterized with FA as fibrovascular pigment epithelial detachment (FV-PED), characterized by a mixture of hyper-fluorescence and irregular elevations, or late leakage of undetermined source (LLUS), characterized by leakage in the late phase . Due to the obscuration of the CNV by rapid dye pooling, the location and exact boundaries of occult CNV often are difficult to determine. Indocyanine green (ICG) angiography can be valuable in detecting and improving the delineation of classic and occult lesions and, in some instances, more precisely guide the treatment. The occult subtype is generally less visible on fundus examinations and is historically less responsive to anti-vascular endothelial growth factor (anti-VEGF) injections, while the classic form is easier to identify and generally has a robust response to anti-VEGF injections .
Focal argon laser photocoagulation could reduce the overall level of vision loss with nvAMD, but its utility was hampered by an immediate, permanent scotoma, along with a high recurrence rate . In the year 2000, the US FDA approved photodynamic therapy (PDT) with verteporfin, which reduced vision loss and avoided an immediate, dense scotoma . In recent years, verteporfin PDT has become less common due to anti-VEGF injections (ranibizumab, bevacizumab, aflibercept, brolucizumab, and most recently faricimab-svoa) which are currently considered gold standard therapy . However, anti-VEGF injections have some limitations, including a significant injection burden (as often as every four weeks), under treatment due to systemic health issues, and a suboptimal treatment response (i.e., persistent or new fluid and/or hemorrhage, progressive lesion fibrosis) that leads to suboptimal vision recovery . Rarely, intravitreal injectables can also be associated with adverse events, including endophthalmitis and retinal detachment . We sought to review the role of PDT as an adjuvant treatment for nvAMD in the current era of anti-VEGF injections. Using the available literature, we compared anti-VEGF injections alone versus in combination with verteporfin PDT for 12-month best corrected visual acuity (BCVA), central retinal thickness (CRT), and injection burden. Supplemental analyses centered around comparing PDT to placebo, and PDT to PDT in combination with intravitreal triamcinolone (IVTA). IVTA has been used as an adjunct to PDT and has characteristics that target the pathophysiology of CNV, and therefore was included to evaluate its role in the treatment of nvAMD.
2
Methods
2.1
Source selection
A systematic search was conducted using PubMed, Cochrane Library, and ClinicalTrials.gov using keywords “macular degeneration” and “photodynamic therapy” (both standard, 50 j/cm 2 and half-fluence, 25 j/cm 2 , were included) and “placebo” or “ranibizumab” or “bevacizumab” or “aflibercept” (2 mg) or “faricimab-svoa” or “brolucizumab” from inception to September 2023. Criteria for inclusion included (1) peer-reviewed retrospective, prospective, or observational primary data studies (including randomized control trials and pilot clinical trials), (2) patients with active CNV secondary to nvAMD, (3) studies reporting results of PDT monotherapy or combination with anti-VEGF and/or intravitreal triamcinolone (IVTA), and (4) studies reporting data on patient demographics, visual acuity (VA), and central retinal thickness at baseline and endpoint (12 months). Studies reporting data on patients with Polypoidal Choroidal Vasculopathy (PCV) were included in data collection, with the goal of a subgroup analysis compared to nvAMD alone. Studies that did not fulfill these criteria, such as reviews, other meta-analyses, case reports, non-human studies, duplicate publications, and single-group pilot studies were excluded. This protocol was registered on the PROSPERO database before data collection commenced.
The initial search yielded 543 total results for screening. Initial screening showed 174 repeat results, 229 studies with unrelated content, 2 meta-analyses, 8 case reports, 1 non-human study, 74 reviews, 8 studies with no comparison group, and 28 overall relevant works. After downloading and completing a full-text review, 5 further studies were excluded for not reporting 12-month follow-up data. 23 studies remained for final meta-analysis. No studies were found comparing faricimab-svoa or brolucizumab and PDT. Preferred Reporting Items for Systemic Reviews and Meta-analyses (PRISMA) guidelines were followed for the selection and sorting of data, shown in Fig. 1 . Where applicable, the research was approved by the institutional human experimentation committee or institutional review board (IRB).
Two independent researchers (SS and KS) completed the Risk of Bias Assessment using the Critical Appraisal Skills Programme (CASP) checklist. An independent third-party researcher reviewed and discussed any discrepancies.
2.2
Data extraction
Ultimately, 2781 patients (2781 eyes) remained for final analysis. All patients had nvAMD at baseline. Two studies reported data on patients diagnosed with PCV; thus, subgroup analysis was not possible. Baseline demographic data included the number of patients in each group, age, visual acuity, and number of patients with previous treatment. Collected variables, when available, were BCVA (at baseline and endpoint), CRT (at baseline and endpoint), and number of total intravitreal injections over 12 months. Rates of cataract formation at 12 months were also recorded for the PDT in combination with IVTA versus the PDT alone group.
2.3
Data analysis
Patients were split into five groups: PDT + intravitreal bevacizumab (IVB) versus IVB alone (group 1), PDT + intravitreal ranibizumab (IVR) versus IVR alone (group 2), and PDT + intravitreal aflibercept (IVA) versus IVA alone (group 3), PDT versus placebo (group 4), PDT alone versus PDT + IVTA (group 5). Group 3 data was found to be insufficient for central analysis as one study reported baseline data with necessary statistical markers (standard deviation) but did not report these necessary outcomes at 12-month. Another study did not separate PDT + IVA versus IVA groups at 12-month outcomes. Attempts to gain this information through corresponding authors were unsuccessful. Therefore, analyses including Group 3 were reported as supplemental. Groups 4 and 5 results were also reported as supplemental analyses as they do not represent novel findings and were accessories to this study’s central aim. Continuous variables of interest were analyzed using fixed effects in the meta-analysis, yielding mean difference (MD), while dichotomous predictors were evaluated with odds ratios using Review Manager 5.4. Anti-VEGF injection monotherapy versus in combination with PDT were pooled for analysis and evaluated as subgroups. Both pooled and subgroup findings were presented in the results for each study variable. Heterogeneity was evaluated by visual examination of funnel plots of I 2 statistics. P value < 0.05 indicated statistical significance.
3
Results
3.1
Patient Characteristics
Ultimately, 2781 patients from 23 studies were included in the final analysis. There were 138 patients in group 1 (PDT + IVB vs IVB), 884 in group 2 (PDT + IVR vs IVR), 159 in group 3 (PDT + IVA vs IVA), 1391 patients in group 4 (PDT vs placebo), and 209 patients in group 5 (PDT vs PDT + IVTA). All patients in group 3 (PDT + IVA vs IVA) had a diagnosis of PCV, a subgroup of nvAMD. Baseline demographic data in Table 1 includes age, baseline BCVA (LogMAR), prior treatment status, and intervention group. In studies that reported data on previous treatments, 97.14 % of patients were treatment-naïve at the beginning of the study.
| Patients | Age | Baseline VA (logMAR) | Treatment-Naive | Prior PDT | Unknown Prior Tx | Intervention | |
|---|---|---|---|---|---|---|---|
| VIO, 2009 | 364 | 79 | 0.57 | 354 | 10 | 0 | PDT v Placebo |
| VIM, 2005 | 79 | 77.5 | 0.61 | – | – | 79 | PDT v Placebo |
| TAP, 1999 | 609 | 75.5 | 0.64 | – | 82 | 527 | PDT v Placebo |
| VIP, 2001 | 339 | 75 | 0.39 | – | 21 | 54 | PDT v Placebo |
| Chaudhary, 2007 | 30 | 79.8 | 0.72 | – | – | 30 | PDT + IVTA v PDT |
| Ruiz-Moreno, 2005 | 45 | 76.2 | 0.75 | – | – | 45 | PDT + IVTA v PDT |
| Arias, 2006 | 61 | 76.7 | 1.01 | – | – | 61 | PDT + IVTA v PDT |
| Hara, 2009 | 73 | 73 | 0.67 | – | – | 73 | PDT + IVTA v PDT |
| Saviano, 2016 | 62 | 78 | 0.60 | 51 | 11 | 0 | PDT + BEVA V BEVA |
| Costagliola, 2015 | 76 | 64.3 | 0.65 | – | – | 76 | PDT + BEVA V BEVA |
| Westhouse, 2012 | 20 | 76.6 | NR | – | – | 20 | PDT + RANI V RANI |
| Larsen, 2012 | 255 | 76.2 | 0.60 | 255 | 0 | 0 | PDT + RANI V RANI |
| Kaiser, 2012 (DENALI) | 216 | 54.2 | 0.61 | 216 | 0 | 0 | PDT + RANI V RANI |
| Krebs, 2013 | 44 | 78.9 | 0.64 | 44 | 0 | 0 | PDT + RANI V RANI |
| Hatz, 2015 | 35 | 78.5 | 0.65 | 24 | 11 | 0 | PDT + RANI V RANI |
| Weingessel, 2014 | 30 | 82.2 | 0.55 | – | – | 30 | PDT + RANI V RANI |
| Bashshur, 2011 | 30 | 73.3 | 0.62 | 26 | 4 | 0 | PDT + RANI V RANI |
| Chen, 2020 | 130 | 59.1 | 0.90 | 130 | 0 | 0 | PDT + RANI V RANI |
| Semerano, 2015 | 50 | 76.9 | 0.60 | 50 | 0 | 0 | PDT + RANI V RANI |
| Williams, 2012 | 56 | 79.2 | 0.67 | 56 | 0 | 0 | PDT + RANI V RANI |
| Vallance, 2010 | 18 | NR | 0.65 | 18 | 0 | 0 | PDT + RANI V RANI |
| Silva, 2020 | 50 | 73.5 | 0.38 | – | – | 50 | PDT + AFLIB V AFLIB |
| Miyakubo, 2023 | 109 | 74.2 | 0.29 | – | – | 109 | PDT + AFLIB V AFLIB |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
