The Efficacy of Conbercept in Polypoidal Choroidal Vasculopathy: A Systematic Review

Methods Thirty studies with 1308 eyes were identified and included in this study. The primary outcome measures were best-corrected visual acuity (BCVA), and secondary outcomes were optical coherence tomography characteristics and polyp regression rates. The pooled results were calculated by the random-effect or fixed-effect model according to the heterogeneity of the data. Results Despite a large standard deviation in means (SMD) improvement for BCVA and central retinal thickness (CRT) in the conbercept group, there was no statistically significant difference in the other outcomes compared to ranibizumab and aflibercept. However, there was a greater polyp regression rate in the conbercept group at 12 months. Conclusions This systematic review indicates that conbercept may achieve similar BCVA and CRT improvements as ranibizumab and aflibercept, with a superior rate of polyp regression at 12 months.


Introduction
Polypoidal choroidal vasculopathy (PCV) is a subtype of neovascular age-related macular degeneration (nAMD) defined by orange-red bulb-like lesions and characteristic polypoidal features with or without a branching vascular network (BVN) on imaging [1,2]. It was first described as posterior uveal bleeding syndrome [3] later described by Yannuzzi et al. as a choroidal vasculopathy which led to hemorrhage and exudation in 1990 [4]. It differs from typical choroidal neovascularization (CNV) in nAMD by the presence of serous or hemorrhagic pigment epithelial detachment (PED) which can cause serous retinal detachment or submacular hemorrhage leading to severe visual deterioration [5].
Antivascular endothelial growth factor (VEGF) therapy, with or without verteporfin photodynamic therapy (vPDT), has been widely used in PCV since randomized clinical trials demonstrated excellent efficacy [6,7]. Several studies have shown that anti-VEGF monotherapy can stabilize the vision and decrease the exudation in PCV, but it had a limited impact on the polypoidal lesions, especially when using ranibizumab [8]. However, it is critical and common to use the polyp regression rate as a parameter in the studies to evaluate the outcome of PCV [7,9]. e anti-VEGF medicines being used now are ranibizumab, bevacizumab, aflibercept, and conbercept. Ranibizumab is a humanized, affinity-matured, Fab fragment against all isoforms of VEGF-A that has been widely used in pivotal clinical trials and has solid evidence indicating improved best-corrected visual acuity (BCVA) and good safety in the management of PCV [10]. Bevacizumab is humanized monoclonal antibody against all isoforms of VEGF-A that has less clinical trial data but has been used successfully in PCV. Aflibercept and conbercept are recombinant fusion proteins composed of key domains from VEGF receptors 1 and 2 and are the only agents that bind all isoforms of VEGF-A, VEGF-B, and placental growth factor. e difference in the molecular structure is that conbercept has an additional domain 4 from VEGF receptor 2 which improves its binding affinity, isoelectric point, and half-life. Conbercept is currently FDA approved in China and undergoing worldwide phase 3 studies in nAMD [11], it has been proved to be safe and effective in nAMD in Chinese clinical trials [12] and has a lower price compared to the other approved anti-VEGF medicines in China.
Some meta-analyses have compared treatment options of PCV [13][14][15]. However, none of them have included conbercept. As a promising new treatment of PCV, it is important to evaluate the efficacy of conbercept in PCV, which will improve our understanding in the treatment options in PCV and optimize the management for PCV patients.

Inclusion and Exclusion
Criteria. Studies were included if they met the following criteria: (1) there were clinical studies on human beings, (2) treatment included monotherapy with conbercept, ranibizumab, or aflibercept for the management of PCV, with or without placebo/control treatment, (3) proportion of VA improvements (visual gain more than 15 letters), the changes of BCVA, and central retinal thickness (CRT) as well as their standard deviation were available, and (4) polyp regression rates were available. Articles without full text, meeting reports, case reports, and repeat publications, were excluded from the analysis.

Data Extraction.
e data was extracted in accordance with the following criterion: (1) the title, corresponding author, year of publication, the number of eyes, average age of patients, follow-up visits, gender composition, interventions, and retreatment regimen, (2) the mean BCVA and its standard deviation (SD) at baseline, (3) the mean BCVA and its standard deviation at month 12, (4) the mean CRT and SD at baseline, (5) the mean CRT and SD at month 12, (6) the percentage of polyp regression at month 12, (7) the mean injection numbers and SD at 12 months, and (8) the frequency of adverse effects, including subretinal hemorrhage and vitreous hemorrhage, which were recorded.

Outcome
Measures. BCVA is a good indicator of retinal function. All visual acuity data were recorded and transformed into logMAR notation for statistical analysis. e CRT thickness was measured by ocular coherence tomography (OCT). e complete polyp regression rate was identified by ICGA. Month 12 was chosen as the time points for evaluating the effects of anti-VEGF treatment in PCV.

Quality Assessment.
According to the Cochrane Handbook for Systematic Reviews of Interventions, we use GRADEpro software (GRADEproGDT 2015) to create the table of summary of find (SOF) ( Table 1). BCVA and CRT at 12 months were included, as well as the polyp regression rate at 12 months. e quality of the evidence was judged using five aspects including limitation in study design, inconsistency, indirectness, imprecision, and reporting bias. e reasons why the quality of the studies was up-or downgraded are listed in the footnotes.
2.6. Statistical Analysis. Extracted data were analyzed by Comprehensive Meta-Analysis version 2 software. Standard deviation in means (SMD) with a 95% confidence interval was used for continuous variables, standard error (SEM), and variance were also calculated. We used fixed-effects model when there is no statistical heterogeneity (I 2 > 50% or p < 0.1); otherwise, the random-effects model was selected. e significance of two anti-VEGFs was judged by p value. A p value <0.05 was considered significant, while p values >0.05 meant the two drugs performed equally.

Search Results.
A total of 833 articles were displayed after searching the keywords (Figure 1). After applying the inclusion and exclusion criteria, 30 studies with 1308 eyes were identified and included in the analysis. ere are 4 arms for 0.5 mg/2.0 mg conbercept, 12 arms for 0.5 mg/2.0 mg ranibizumab, and 14 arms for 2.0 mg aflibercept. In studies comparing anti-VEGF monotherapy and PDT monotherapy, only data related to anti-VEGF monotherapy was included in our study.

Assessment of Evidence Quality.
We used the "Grades of Recommendations Assessment, Development and Evaluation" (GRADE) system for evaluating the quality of evidence. e GRADE system would upgrade or downgrade the recommendation intensity according to several standards. As shown in Table 1, the GRADE of comparison of polyp regression rate between conbercept and ranibizumab at 12 months was low, while the others were very low.
Egger's measures were conducted to measure the asymmetry. e p value of Egger's measure of two outcomes "polyp regression rates" and "change in BCVA," are higher than 0.05 (0.456 in polyp regression rates and 0.226 in changes of BCVA). However, p-Egger's of "change in CRT" s is 0.01.

Study Characteristics.
e baseline characteristics of the analysis are shown in Table 2. e duration of the studies varied from 1 month to 36 months. All the studies included the BCVA improvement and CRT change. e rates of polyp regression identified by ICGA were contained in the studies.

Secondary Outcome: Change in CRT.
e pooled-analysis of CRT results is shown in Table 4. At month 12, conbercept (p < 0.0001, 95% CI: 0.690 to 1.476), ranibizumab (p < 0.0001, 95% CI: 0.688 to 0.950), and aflibercept (p < 0.0001, 95% CI: 1.098 to 1.772) showed a significant decrease in CRT based on OCT imaging. Similar to the change in BCVA, there was no significant difference between conbercept and the other two anti-VEGF medicines in terms of CRT (aflibercept: p � 0.331, 95% CI: 1.063 to 1.610) (ranibizumab: p � 0.144, 95% CI: 0.722 to 1.114).   3 Moderate * e basis for the assumed risk (e.g., the median control group risk across studies) is provided in footnotes. e corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval. GRADE working group grades of evidence. High quality: further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: we are very uncertain about the estimate. 1 Heteorgeneity. 2 Indirect comparation. 3 No explanation was provided. 4 Different dose in conbercept group.

Journal of Ophthalmology
Interestingly, there is a significant improvement in CRT in the aflibercept group than the ranibizumab group (p � 0.032, 95% CI: 0.976 to 1.473) ( Table 3).

Discussion
PCV is a subtype of nAMD with characteristic clinical and imaging features that differentiate it from typical nAMD. It has been reported that the levels of VEGF are lower in the PCV eyes than typical nAMD eyes [45], possibly indicating a different underlying pathophysiology as well as treatment response to anti-VEGF therapy. Studies such as the PEARL study [46] (PCV management with different concentrations of intravitreal ranibizumab (IVR) monotherapy) and the LAP-TOP study [47] (comparison of PDT versus IVR in PCV) have shown excellent efficacy of anti-VEGF agents in PCV. More recently, the EVEREST 2 and PLANET studies verified that anti-VEGF monotherapy can effectively treat PCV [48][49][50].  Journal of Ophthalmology ere have been numerous comparative studies of different anti-VEGF agents in a head-to-head manner in typical nAMD, but to date, no comparison studies in PCV management have occurred [8]. As a newly developed anti-VEGF medicine, limited studies focused on the efficacy of conbercept in comparison to ranibizumab and aflibercept in PCV, despite that conbercept is superior to ranibizumab in improving visual outcome in typical nAMD patients [51].
We performed this systematic review of published clinical studies evaluating conbercept and other anti-VEGF agents in the management of PCV. It showed significant improvement in BCVA and reduction in CRT among the three anti-VEGF agents at 12 months. In addition, conbercept showed the significantly highest rate of polyp regression at 12 months and less injection frequency than aflibercept.
e conbercept and ranibizumab group are comparable since all of them were using a 3 + PRN (additional injections were given as needed after 3 initial monthly dose) regimen. However, almost half of the studies included for the injection frequency analysis in the aflibercept group were using a regimen that received every 2 months after 3 initial monthly doses, and one study is treatand-extend regimen, while the rest were using a 3 + PRN regimen. erefore, we want to remind readers to interpret these results with caution. As a structurally similar anti-VEGF to conbercept, aflibercept monotherapy has also demonstrated ideal efficacy in both PLANET [50] and VAULT study [43]. However, conbercept possesses better binding affinity and half-life due to its additional domain 4 from VEGFR2, which may improve its efficacy at 12 months.
CRT, a frequently used indicator to describe anatomic outcomes in nAMD studies [52], can also be used to evaluate treatment response in PCV. All the anti-VEGF agents in our study demonstrated significant decreases in CRT at month  12. Aflibercept has a slight advantage over ranibizumab, while there is no difference was found between aflibercept and conbercept. ese results may indicate that there is a potential advantage of fusion proteins over ranibizumab on decreasing exudation and polyps. One of the common parameters in PCV is polyp regression. Polyp regression rate is a crucial factor to the visual outcome in PCV patients, since the exudation and rupture of polyps would cause the enlargement of PED as well as massive hemorrhage [9,53]. us, we looked at polyp regression rates in this meta-analysis. Conbercept demonstrated a significantly higher rate of polyp regression (with a mean event rate of 0.683) than ranibizumab (an event rate of 0.324) and aflibercept (an event rate of 0.496) at 12 months. Aflibercept monotherapy was reported to have a 38.9% complete polypoidal regression rate at week 52 in the PLANET study [50]. e polyp regression rate was 33.8% in EVERESTII (ranibizumab monotherapy) [49] and 56.5% in the subgroup analysis of the AURORA study (conbercept monotherapy) [16]. Combining our results, it seems that conbercept has an advantage over aflibercept and ranibizumab in the closure of polyps, though the methodological differences among studies must be considered.
A possible explanation for the higher polyp regression rate of conbercept may be related to its high binding affinity for VEGF-A. By combining the lg-like domain of VEGFR1 and VEGFR2, it has a higher binding affinity for VEGF-A 165 and lower isoelectric point than ranibizumab [54,55]. Moreover, by design, conbercept binds placental growth factor (PGF) and VEGF-B.
is approach is similar to aflibercept, another anti-VEGF fusion protein, that also has reported a higher rate of polyp regression than ranibizumab and bevacizumab [33,42]. Aflibercept has a similar structure to conbercept, but with a lower binding affinity for VEGF-A, VEGF-B, and PGF. Moreover, we noticed that aflibercept had a significant decrease in CRT compared to ranibizumab, while conbercept did not. It may be due to the small quantity of study numbers in conbercept group and brings the wide range of confidence interval. e PANDA study (NCT03630952 and NCT03577899) is an international trial comparing conbercept with aflibercept, which will provide better data for this comparison of decrease in CRT with anti-VEGF therapy.
ere are some limitations in our study. First, although we searched as many websites and databases as possible in order to include all the present studies, we may still have missed some papers, resulting in the publication bias. Second, since there was only a small quantity of randomized controlled trials (RCTs) with conbercept, most of the studies in conbercept group were retrospective which may lead to the ascertainment bias.
us, we had to compare conbercept, aflibercept, and ranibizumab indirectly. ese biases were downgraded in the SOF table, and our results were regarded as low-quality in GRADE system. e current definition of low quality evidence is that the confidence in the effect estimate is limited, but not worthless. Very low quality evidence stands for less confidence in the effect estimate. erefore, we want to remind readers to interpret the results with caution. Obviously, high-quality RCTs and additional detailed data would help eliminate these biases in the future.
To our knowledge, this is the first systematic assessment of the efficacy of conbercept in PCV, using ranibizumab and aflibercept as references. Our study demonstrated that conbercept can achieve similar BCVA and CRT outcomes as ranibizumab and aflibercept during the first 12 months of treatment. Moreover, conbercept was superior to ranibizumab and aflibercept in outcome of polyp regression at one year. e data from ongoing randomized clinical trials will provide more details comparing anti-VEGF response between ranibizumab, aflibercept, and conbercept for the management of PCV.

Disclosure
Yimin Wang and Mengxi Shen are the co-first authors.

Conflicts of Interest
PKK is a consultant at Allergan, Bayer, Kanghong, Kodiak, Novartis, and Regeneron. XDS is a consultant at Allergan, Bayer, Kanghong, Kodiak, and Novartis. e authors declare that they have no conflicts of interest.