Five-year real-world outcomes of anti-vascular endothelial growth factor monotherapy versus combination therapy for polypoidal choroidal vasculopathy in a Chinese population: a retrospective study

Polypoidal choroidal vasculopathy (PCV) has been recognized as a major cause of significant and acquired vision loss in elderly patients [1‐3]. PCV is characterized by polypoidal or aneurysmal hyperfluorescence on indocyanine green angiography (ICGA) [4, 5]. Branching vascular network (BVN) and serosanguinous pigment epithelium detachment are frequently associated with polypoidal lesions [6, 7]. PCV is a subtype of neovascular age-related macular degeneration (nvAMD), and PCV also falls in the pachychoroid spectrum diseases [6].

Although its pathogenesis is still unknown, anti-vascular endothelial growth factor (VEGF) and photodynamic therapy (PDT) have been used for treating PCV in real world. Unlike typical nvAMD, where intravitreal anti-VEGF therapy has been the mainstay of treatment for over a decade, currently anti-VEGF therapy, verteporfin PDT, and various combinations of these therapies for PCV exist simultaneously [8‐12]. In order to achieve the best possible visual outcomes, both EVEREST-II and PLANET studies indicated that anti-VEGF therapy and combination therapy with PDT give excellent functional visual outcomes at 1 year [9, 12]. Although long-term favorable visual outcome with PDT and anti-VEGF therapy has been validated respectively, [13‐19] little is known about that, among anti-VEGF monotherapy and various combination of PDT and anti-VEGF therapy, which treatment is superior with respect to long-term best-corrected visual acuity (BCVA) outcome in real world. Regression of polyps is a significant aspect of treatment, nevertheless identifying the treatment that is superior for visual outcome was regarded more important to patients than polyp regression [6, 7, 10].

In this study, we evaluated long-term visual outcomes at least 5 years after anti-VEGF monotherapy and combination of PDT and anti-VEGF therapy for various subtypes of PCV in real world. We also investigated imaging morphological features associated with long-term visual outcomes in these patients.

In total, 53 eyes of 46 patients who had completed at least 60 months of follow-up visits after initial treatment were enrolled. Patients had a mean age of 63.30 ± 10.10 years and showed a male predominance (34 male, 64.2%). Baseline clinical details are listed in Table 1. Twenty-eight eyes received anti-VEGF monotherapy, and 25 eyes received combination therapy, in which 14 eyes received initial combination therapy, and 11 eyes received deferred combination therapy. There were no significant differences of baseline clinical characteristics among three treatment regimens except that the anti-VEGF monotherapy group had greater proportion of presence of subretinal fluid than the initial combination group did significantly (21/28 vs 4/10, P = 0.012, Chi-square test) (Additional file 1: Table S1).

In this retrospective study, we compared the five-year outcomes of anti-VEGF monotherapy and combination therapy for various subtypes of PCV. Fifty-three treatment-naïve PCV eyes treated with anti-VEGF monotherapy and combination therapy were analyzed. The subtype 2 PCV had a favorable visual outcome, while the subtype 1 PCV had a relatively unfavorable visual outcome. Anti-VEGF monotherapy and initial combination therapy achieved more favorable visual outcome compared with deferred combination therapy, especially for subtype 2 PCV. Initial combination therapy needed less number of PDT than deferred combination therapy. Imaging morphological features, including the continuity of RPE at baseline, were associated with the decision of treatment and the visual outcomes at months 60, which could be regarded as imaging biomarkers that might predict response to therapy.

We compared anti-VEGF monotherapy and combination therapy, which were recommended by recent guideline and studies [6, 7, 9, 12]. In 2 recent randomized controlled trial (RCT), EVERST-II and PLANET studies which used different anti-VEGF agents, both therapy regimens with 12-month duration were compared [9, 12]. EVERST-II study demonstrated that combination therapy had superior visual and angiographic outcomes, and a less number of anti-VEGF therapies [9]. However, no comparison on long-term outcomes of anti-VEGF therapy and combination therapy was performed. Additionally, the patients in EVEREST-II study was assigned into various groups randomly, which is scarcely possible in real world because differences in individual characteristics of PCV eyes and financial burdens should be considered when deciding on the best treatment option for each patient. Besides, compared with EVEREST-II study, PLANET study concludes that no additional benefit in combining with PDT as a rescue therapy, which suggested that the outcome of initial combination regimen of anti-VEGF therapy and PDT might differ from that of deferred combination regimen of anti-VEGF agents plus deferred PDT [12]. Similarly, the combination therapy in the present study was divided into 2 groups as well: the initial combination and deferred combination. When comparing these treatment regimens, no significant differences of baseline clinical features were detected except subretinal fluid, which was not considered in decision on various treatment regimens and has not been proved, as a structural feature, to be related to long-term visual acuity outcomes directly.

Several studies have reported various long-term visual outcomes of PCV treated with various regimens. Saito et al. retrospectively evaluated the 60-month outcome of PDT with or without anti-VEGF therapy for PCV in 60 eyes [18]. The mean logMAR visual acuity at baseline was 0.66 and after 60 months was 0.71. Kang et al. reported an improved visual outcome from a mean baseline BCVA of 0.78 logMAR to the 5-year BCVA of 0.67 logMAR in 42 PCV eyes treated with PDT [17]. On the contrary, unfavorable visual acuity was also detected in another study reported Chang et al. They retrospectively evaluated the 4-year treatment outcome of 31 PCV eyes treated with anti-VEGF therapy [15]. The mean visual acuity at baseline was 0.52 logMAR and after 4 years was 0.83 logMAR. However, all the previous studies did not classify enrolled eyes into various subtypes. In the present study, we classified the enrolled eyes into 2 subtypes according to criteria provided by previous studies [20‐22]. The eyes of subtype 1, polypoidal CNV, had an unfavorable visual outcome, which deteriorated from 0.69 to 1.25 logMAR significantly, while the eyes of subtype 2, typical PCV, had a favorable visual outcome, which improve from 0.62 to 0.57 logMAR without significant difference. Jang et al. also reported a better BCVA at baseline and at months 12 after the initial treatment in eyes with subtype 2 PCV than that in eyes with subtype 1 PCV [20]. The present study revealed a long-term course of treatment on PCV, and suggested that various subtypes of PCV in real world presented totally different long-term visual outcomes.

The mean number of anti-VEGF therapies in the anti-VEGF monotherapy group was 4.64, which was less than that of previous studies. An extensive study of LAPTOP study for totally 5 years reported a mean injection number of 14.8 for ranibizumab in the anti-VEGF monotherapy group, in which the number of ranibizumab injection was 8.0 in the last 3 years after the LAPTOP study, while the number of aflibercept was 3.7 [23]. The injection number of conversive therapy was less than the injection number of only ranibizumab. Another Japanese study reported a mean injection number of 18.2 for ranibizumab [19]. The mean injection number of anti-VEGF monotherapy in the present study was much less than the previous studies. Conversive therapy might help reduce the injection number. Besides, eyes with less severity and activity was treated with anti-VEGF monotherapy rather than combination therapy in real world according to guidelines, [7] which might lead to a less number of injections. Additionally, in order to reduce the selection bias due to the limitations of retrospective studies, not only eyes needed persistent treatment were enrolled, eyes with favorable visual outcomes were also enrolled, which might contribute the less injection number. Furthermore, the enrolled Chinese patients might bear the financial burden, and the anti-VEGF agents used in this study have not been paid by medical insurance during majority of the follow-up period, which indeed reduced the patients’ therapeutic compliance. What’s more, the retreatment criteria in the present study was not totally same as that of previous studies, [18, 19, 23] and the treatment decision were at the discretion of the treating physicians rather than a strict prospective protocol. And some patients with lesions of less activity and severity were allowed to visit us every 6 months. For majority eyes enrolled in the present study, only explicit signs of recurrence rather than potential signs were present, retreatment would be performed on the eyes showing response to treatment. Therefore, comparing with previous studies, patients might receive anti-VEGF therapies of less injection numbers in real world, especially in developing countries.

The present study revealed differences on the changes of BCVA at various time points and treatment number between initial and deferred combination regimen, which suggested that they should be analyzed respectively. PDT has been proved to be effective in the management of PCV [7]. And in real world, initial PDT with deferred anti-VEGF therapy has been performed. However, its long-term outcomes have not been investigated. In the present study, the initial combination group requires a significantly less number of PDT than the deferred combination group. When considering the similar BCVA at baseline, the initial combination regimen had a more favorable BCVA outcome at months 30 in eyes with continuous RPE and in eyes of subtype 2 PCV. And the initial combination regimen was more likely to earn improved and stable BCVA outcomes than deferred combination regimen. Therefore, when combination regimen was needed, initial combination of PDT and anti-VEGF therapy would be recommended by the present study.

Imaging morphological features associated with outcomes were also investigated. Firstly, we divided eyes into 2 subtypes, and we found that the subtype 2 PCV had a much more favorable visual acuity. Secondly, deteriorated BCVA has more percentage of discontinuous RPE. Neurosensory retina might be affected by abnormal vessels or BVN directly when RPE was discontinuous, and the higher percentage of discontinuous RPE might contribute to the deteriorated BCVA outcomes. At the meantime, we found that, the eyes with continuous RPE treated with anti-VEGF monotherapy had a better visual outcome than the eyes treated with deferred combination therapy during the follow-up period. PDT could lead to choriocapillary occlusion, RPE and neuroretina injury [24, 25]. Therefore, we speculated that the dysfunction of RPE and outer layers neuroretina after PDT without anti-VEGF therapy might lead to an unfavorable visual outcome. On the contrary, anti-VEGF agents can alleviate the inflammation reaction, [26, 27] which might protect retina and choriocapillary in the eyes of initial combination group. Thirdly, the formation of polyps was associated with BCVA outcomes. Polyps are a clinical characteristic of PCV, and regression of polyps plays an important role in the management of PCV [7, 10]. Suzuki et al. reported that multiple polyps showed more correlation with poorer BCVA, which suggested that the number of polyps might be valuable to understand the prognosis of PCV [28]. In the presented study, we noticed that the formation of polyps were also associated with visual outcomes. Hou et al. and Cackett et al. summarized 3 formation of polyps, including single, cluster, and string, but the relationship between the formation of polyps and visual outcomes was not investigated [29, 30]. Single polyps referred to isolated polyps, while cluster and string polyps referred to interconnected polyps. Therefore, we summarized the formation of polyps into 2 types, including the isolated and interconnected polyps. In the present study, interconnected polyps occurred more frequently in eyes of deteriorated BCVA. Polyps at baseline may reflect varying activity and severity of PCV, and further study will be needed to definitively determine the prognostic and mechanistic impact of polyps in PCV.

Our study has several limitations, including the relatively small patient number, especially in the initial and deferred combination therapy groups. Given the small numbers in these groups, it is difficult to draw firm conclusions in some situations. Because of the retrospective nature, it was difficult to obtain regular follow-up angiograms which did not affect decisions regarding retreatment. Another possible limitation is selection bias. In order to minimize potential selection, we enrolled eyes treated with different regimens and eyes with various visual outcomes. However, only Chinese patients were enrolled, and worldwide multicenter investigations might be needed to study PCV in real world. Additionally, although no RCTs demonstrated that any anti-VEGF agents were superior to the others, it is unknown the impact of conversion of anti-VEGF agents on the outcomes of PCV in the present study, we could not evaluate the different effects of various anti-VEGF agents on the visual outcomes.

In conclusion, the 5-year outcomes of anti-VEGF monotherapy and combination therapy for PCV were reviewed. Our study demonstrated that anti-VEGF monotherapy and initial combination therapy were favorable overall over a 60-month follow-up period. Eyes with subtype 2 PCV and eyes with continuous RPE or isolated polyps were more likely to gain a favorable visual outcome. Decisions on treatment regimens should take the imaging features of PCV lesions into consideration. Because this study was a retrospective review with limited size, large, long-term, and prospective randomized studies are needed to determine the efficacy and safety profiles for patients with PCV of various severity and activity.