Regression of Neovascularization after Panretinal Photocoagulation Combined with Anti-VEGF Injection for Proliferative Diabetic Retinopathy—A Review

Proliferative diabetic retinopathy (PDR) poses a significant therapeutic problem that often results in severe visual loss. Panretinal photocoagulation (PRP) has long been a mainstay treatment for this condition. Conversely, intravitreal anti-VEGF therapy has served as an alternative treatment for PDR. This review aimed to evaluate the effects of PRP combined with anti-VEGF therapy on the regression of neovascularization (NV), including functional outcomes and incidence of complications. The MEDLINE database was searched for articles evaluating regression of NV using a combination of the following terms: “proliferative diabetic retinopathy”, “anti-VEGF”, “panretinal photocoagulation”, and “combined treatment”. The search yielded a total of 22 articles. The analysis of their results indicated PRP combined with ant-VEGF therapy as superior over PRP alone in the management of PDR. Combination treatment yields better and faster regression of NV and a lower incidence of serious complications, such as vitreous hemorrhage and the need for pars plana vitrectomy. Nevertheless, complete regression of NV is not achieved in a significant proportion of patients. Further research is needed to establish the most effective schedule for intravitreal injections as an adjunct to PRP. The current literature shows that in some cases, cessation of anti-VEGF injection in combination treatment for PDR can lead to relapse of NV.


Introduction
Diabetic retinopathy is one of the leading causes of visual loss in developed countries.Among its subtypes, proliferative diabetic retinopathy (PDR) is considered a vision-threatening diabetic retinopathy, as it is burdened with severe potential complications that affect patients' sight.According to the Wisconsin Epidemiology Study for Diabetic Retinopathy (WESDR), the prevalence of PDR is 23% in patients with an earlier onset of diabetes, 10% in those with a later onset of diabetes, and 3% in those who do not take insulin [1].The 4-year risk of the development of PDR increases with the presence of hyperglycemia, a longer duration of diabetes, and a more severe retinopathy at baseline.In the WESDR, a cumulative 25-year risk of progression to PDR of 42% is reported [2].This finding shows that the total number of patients with PDR is relatively large, indicating the need for close observation and frequent prompt initiation of treatment.The serious complications of PDR include vitreous hemorrhage (VH), vitreoretinal traction, and tractional retinal detachment or retinal tears that can lead to retinal detachment through a rhegmatogenous mechanism, which usually require surgical intervention, including pars plana vitrectomy (PPV).The outcomes of vitreoretinal surgery for advanced forms of PDR are often unsatisfactory.Despite achieving good morphological outcomes, patients often experience some degree of visual impairment [3,4].
Early initiation of PDR treatment can prevent serious complications and result in better visual outcomes in the long term [5,6].Thus, the aim of modern ophthalmology is to provide therapeutic solutions that would minimize the need for surgical intervention and achieve satisfactory functional outcomes.
Panretinal photocoagulation (PRP) has long been the mainstay treatment for PDR.Since the first Diabetic Retinopathy Study Group studies in the 20th century, much quality randomized research has confirmed the efficacy and reliability of PRP [7][8][9].However, the mechanism of retinal ablation via laser photocoagulation remains poorly understood.The most prevalent pathophysiological concept is based on the oxygen theory [10,11].According to this concept, PRP creates scars that serve as bridges for the oxygen flow from the choroid to the inner parts of the retina.Accordingly, oxygenation of the retina significantly improves, reducing the production of vasoproliferative cytokines [12].This effect is enhanced by the reduction of retinal need for oxygen as a consequence of destruction of a large number of photoreceptor cells that normally consume large amounts of oxygen [13].Nonetheless, PRP is a destructive procedure associated with potential complications such as restriction of the visual field, proliferation of glial cells on the retinal surface, formation of epiretinal membranes, and detachment or neovascularization (NV) of the choroid [14].Moreover, the procedure is not 100% effective despite extensive ablation of the peripheral retina.Some studies report that in as much as 20-30% of patients, PDR does not regress or deteriorate despite receiving full laser treatment [15][16][17].
Modern ophthalmology is directed toward retina-sparing forms of treatment, with intraocular injections of anti-VEGF agents considered to be at the first line.Anti-VEGF therapy works by blocking receptors for vasoproliferative cytokines that are elicited by ocular tissues in the state of hypoxia.Consequently, VEGF-related neo-vessel growth is hampered, and VEGF-related hyperpermeability of retinal vessels is reduced.Drugs have been originally formulated for the treatment of macular diseases, including exudative forms of age-related macular degeneration, diabetic macular edema (DME), and macular edema secondary to retinal vein occlusion [18,19].Nevertheless, they have been administered off-label as auxiliary procedures to minimize bleeding during and after PPV, speed up the resorption of intravitreal hemorrhage, or, occasionally, enhance the regression of NV in proliferative retinopathies [20][21][22][23].Despite the marked progress in the evolution of treatment schedules for anti-VEGF drugs, they must still be administered at a relatively high frequency to ensure effectiveness.Moreover, achieving permanent regression of the treated clinical entity only through intravitreal injections and without the need for further treatment remains controversial [24].
As both PRP and intravitreal anti-VEGF therapy target the development of NV, combining them for the treatment of PDR appears intuitive.The potentially additive actions of both therapies could theoretically yield better regression of neo-vessels and prolong the durability of the therapeutic effect without the need for extensive ablation of the peripheral retina.Lighter PRP is associated with visual field preservation and fewer complications, as proven in studies comparing between light and classic retinal photocoagulation and multispot laser treatment [25,26].
The aim of this review was to evaluate the effects of combination treatment of PDR on regression of retinal NV, including functional outcomes and incidence of complications.

Materials and Methods
The MEDLINE and PubMed databases were searched for articles evaluating regression of NV using different combinations of the following terms: "proliferative diabetic retinopathy", "anti-VEGF", "panretinal photocoagulation", and "combined treatment".The identified articles were reviewed according to their methodology.

Results
The database search yielded a total of 34 studies that included combination treatment: anti-VEGF plus PRP in the management of PDR.Among them, 22 analyzed regression of NV and were included in the review.The sample size, follow-up duration, study design, best-corrected visual acuity (BCVA) change, central subfoveal thickness (CST) change, NV regression, reported complications and adverse events, and other data presented in each article are summarized in Table 1.
In 14 articles, the authors compared the efficacy between PRP alone and PRP combined with anti-VEGF therapy, while in three articles, the authors compared the efficacy between anti-VEGF monotherapy and combination treatment.In two articles, three arms were compared: PRP, PRP combined with intravitreal anti-VEGF therapy, and intravitreal anti-VEGF therapy alone.Two articles compared two versions of combination treatments: anti-VEGF therapy with standard Early Treatment Diabetic Retinopathy Study (ETDRS) PRP and anti-VEGF therapy with modified laser treatment.In one article, only combination treatment (IVR plus PRP) results were reported.Below, we present a synthesis of the reported findings based on the analyzed aspect of combination therapy.BCVA-best-corrected visual acuity, CST-central subfoveal thickness, PRP-panretinal photocoagulation, PRI-panretinal photocoagulation at ischemic areas only, ETDRS-Early Treatment Diabetic Retinopathy Study, NV-neovascularization, NVD-neovascularization at disc, NVE-neovascularization elsewhere, DME-diabetic macular edema, FA-fluorescein angiography, VH-vitreous hemorrhage, PDR-proliferative diabetic retinopathy, HR-PDR-high-risk proliferative diabetic retinopathy, PPV-pars plana vitrectomy, DA-disc area, DD-disc diameter, IVB-intravitreal bevacizumab, IVI-intravitreal injection, IVR-intravitreal ranibizumab, IVC-intravitreal conbercept, IOP-intraocular pressure, RCTrandomized controlled trial, PRN-pro re nata (as needed), FAZ-foveal avascular zone, EZ-ellipsoid zone, ELM-external limiting membrane, TRD-tractional retinal detachment, AE-adverse effect, MD-VF-mean deviation of visual field, ERG-electroretinogram, HbA1c-glycated hemoglobin.

NV Regression
In all studies included in the review, significant regression of NV was reported for combination treatment of PDR.The studies comparing PRP alone with PRP plus anti-VEGF therapy showed greater reduction of the NV area with combination therapy [27,28,[30][31][32][33][35][36][37]39,40,44,47].Conversely, in the recent retrospective study by Si et al. [48], such an advantage of combination treatment over PRP alone was not proven within the 24-month follow-up.Regarding the strength and durability of combination treatment in terms of the treatment schedule and duration of follow-up, early studies [27][28][29][30] utilized only a few months of follow-up and one or two intravitreal injections.Despite prominent NV regression in the first month's post-treatment, the effect diminished over time without further management [28,29].Notably, it is essential to review subsequent studies with developed retreatment criteria and long-term follow-up.Figueira et al. [39] conducted a randomized prospective multicenter research (PROTEUS study) comparing PRP plus intravitreal ranibizumab (IVR) with PRP alone within 12 months of follow-up.The study design involved three monthly injections of IVR and allowed repetition of PRP treatment in case of persistent NV.The reduction of the total NV area was significantly greater in the PRP plus group; however, complete NV regression was obtained only in 43.9% of cases in that group.Moreover, the advantage of combination treatment was significant for neovascularization elsewhere (NVE) but not for neovascularization at disc (NVD).The rate of NV relapse after initial improvement was similar in both groups, but NV reoccurred after complete regression only in the combination group.NVD resistance to treatment was also observed by Zhou et al. [36].Combination treatment yielded complete regression of NVE in 100% of cases in the combination group, but complete regression was noted only in 50% of the eyes.
Generally, despite the superiority of combination treatment to PRP in eliciting reduction of NV, total regression of NV was not achieved in 100% of the eyes in only one study.As an adjunct, anti-VEGF therapy improves morphological outcomes but does not provide complete cure.
PRP combined with anti-VEGF therapy and anti-VEGF monotherapy were rarely analyzed among the studies reviewed [38,43].In the study by Messias et al. [38], regression of NV was similar between the IVR monotherapy group and two laser plus IVR groups.However, notably, repeated injections were allowed in cases of incomplete NV regression or DME; thus, the study protocol included no limit for the number of IVR sessions during the study.The study by Chatziralli et al. [43] utilized a similar design and demonstrated more optimistic results.The authors found total NV regression in the combination group (IVR plus PRP) at the end of the 24-month follow-up.Nevertheless, during that period, patients received a mean number of 11 intravitreal injections in the combination group compared with 14 intravitreal injections in the IVR monotherapy group.Hence, it is plausible that discontinuation of intravitreal treatment can result in NV relapse in some cases.
PRP alone, PRP plus anti-VEGF therapy, and anti-VEGF therapy alone were compared in three studies [34,41,46].Figueira et al. [34] did not report any significant difference in the total NV regression between the groups, although the PRP alone group showed poorer NVD regression than did the IVR and IVR plus groups.Notably, the proportion of patients with total NV regression at the end of the study was small in all groups (<50%).Lang et al. [41] reported similar results during 12 months of follow-up in their PRIDE study.Despite significant reduction of the NV area in the IVR and IVR plus groups, complete NV regression at 12 months was noted in a minority of patients in each group.The same authors reported the results of the second year of observation of patients from their PRIDE study, during which patients were treated under real-life conditions, and anti-VEGF medications were rarely administered [49].Most patients received supplementary PRP treatment.As a result, the NV area significantly increased, prompting the authors to conclude that discontinuation of anti-VEGF treatment for PDR might result in increased NV area and visual loss.
Shahraki et al. [46] also reported superiority of combination treatment (IVB plus modern PRP) over IVB alone and PRP alone in reducing the leakage area.The difference was prominent in the pairwise comparison between combination treatment and PRP alone (p = 0.003).The study design allowed rescue IVB sessions in cases without NV regression or in cases of DME.Leakage from NV was reduced but was still noted at the endpoint of the study.These findings indicate that complete NV regression should not be expected after only a limited number of intravitreal injections.
The decision to use combination treatment instead of PRP alone or anti-VEGF monotherapy for PDR must be evaluated in the context of the results of the Protocol S study by the DRCR net group [50].This milestone study favored IVR over PRP in terms of functional outcomes: patients treated with intravitreal injections had a better visual field and lower incidence of DME.However, the long-term observation of patients showed that intravitreal therapy alone for PDR was burdened with a high proportion of patients who were lost to follow-up and eventually developed serious complications, such as retinal detachment and iris NV, which were less frequent in patients treated with PRP alone [24].Hence, patient compliance is a crucial factor that determines the long-term outcomes of patients treated with anti-VEGF therapy alone [51,52].Moreover, the cost utility of intravitreal monotherapy for PDR is less favorable than that of PRP.The favorable results of combination treatment for PDR suggest its potential superiority over IVR alone and PRP alone.

Number of Required Laser Spots/Energy
In most studies reviewed, the standard ETDRS protocol for PRP was applied, but modified protocols were also utilized and evaluated [38,42,45,46].Two of these studies compared the effects between PASCAL laser combined with anti-VEGF therapy and classic ETDRS laser combined with the same anti-VEGF therapy [38,42].Both lasers proved to be equally effective in reducing NV leakage.Shahraki et al. [46] evaluated a modified PRP protocol that involved PRP anterior to the equator only in combination with IVB.This treatment proved to be superior to PRP alone and IVB alone in causing regression of NV at the end of 12-month follow-up.Toscano et al. [45] compared classic PRP plus IVR with PRP at ischemic areas only plus IVR and found no difference in the outcome between both treatments.
Messias et al. [33], Figueira et al. [39], Yan et al. [35], and Sun and Qi [47] reported that combination treatment reduced the number of laser spots or energy compared with PRP alone.
As shown in the analyses of the results of the reviewed studies, retina-sparing techniques, including reduction of the number of ETDRS PRP spots, multispot PASCAL laser (number of produced spots is generally smaller), PRP at ischemic areas only, and PRP limited to the area anterior to the equator in combination with anti-VEGF therapy were equally as effective as protocols involving a traditionally larger retinal ablation.To date, the number of available studies on the subject is limited, indicating the need for further research.Nevertheless, existing evidence shows that intravitreal therapy as an adjunct to peripheral laser might save some function of the retina owing to a smaller number of required laser spots.

Treatment Schedule
The analyses of the treatment schedules applied in the studies reviewed showed a variety of approaches used and a lack of precise recommendations for combination treatment of PDR.The studies with short follow-up (1-6 months) applied a limited number (one to two) of intravitreal injections in combination with PRP at the beginning of the study and sometimes after the completion of PRP.Such approach was burdened with the loss of the effect of treatment and relapse of NV, as mentioned earlier [28,29].The long-term follow-up studies used more than one or two intravitreal injections in cases without NV regression or in cases of NV relapse throughout the study.In all these studies, a necessity for the continuation of intravitreal treatment after the application of initial doses was indicated [41][42][43][45][46][47][48].These findings indicate that total regression of NV cannot be obtained with only a small number of anti-VEGF injections.The number of required intravitreal treatments is similar for combination therapy and intravitreal therapy alone [34,38,41].Only one study reported fewer injections for combination therapy [45].

Adverse Events: VH and Need for PPV
VH and subsequent treatment with PPV during follow-up were not frequently reported in any study.Moreover, the incidence of these adverse events was not analyzed in all studies.Nevertheless, the incidence of these complications in most studies was higher in the PRP alone group than in the PRP plus group [30,31,34,36,39,46,47].For example, Figueira et al. [34] reported an incidence of 30.8% and 9% in these groups, respectively, while Sun and Qi [47] noted these complications in 27 and seven eyes in their sample of 165 eyes, respectively.These findings are consistent with those of the Protocol S study, which compared anti-VEGF treatment with PRP for PDR.In the Protocol S study, the need for PPV was higher in the PRP group: 15% vs. 4% [50].Nevertheless, the protective effect of anti-VEGF treatment depends on regular application of intravitreal injections.

BCVA Change
The BCVA is expected to improve in patients with PDR with concomitant DME.The reviewed studies often presented results for the whole cohort of patients with PDR, without division to those with and without DME, making it difficult to analyze the real impact of treatment modalities on the BCVA.Accordingly, it is essential to review studies excluding patients with PDR with concomitant DME from the analysis.Among the analyzed studies, four specifically included patients with PDR but without DME at baseline [27,30,39,41].In two of these studies [27,30], no significant difference was noted in the final BCVA between the PRP alone group and PRP combined with anti-VEGF therapy group.No improvement of vision was also observed.In the PROTEUS study by Figueira et al. [39], the BCVA was better in the IVR plus group than in the PRP alone group only after correction for age.Nevertheless, the mean BCVA slightly declined in both groups.The PRIDE study by Lang et al. [41] revealed only a minor improvement in the BCVA of the IVR alone group and a decline in the BCVA of the PRP plus and PRP alone groups.A significant difference was noted only between the IVR and PRP alone groups.These findings indicate a tendency for sustainability of the BCVA with anti-VEGF therapy as an adjunct to PRP.PRP alone is usually associated with BCVA decline.
The composition of the study groups that included patients with DME is similar between the different arms of the reviewed studies (the same proportion of patients with DME).Thus, some conclusions can be drawn.The studies that included patients with DME generally showed either improvements in the BCVA [28,29,32,35,37,38,40,[42][43][44]47] or stabilization [31,41,45] with combination treatment and, consequently, decline or rare stabilization with PRP alone.In the long-term, the effect was transient when anti-VEGF therapy was applied only at the beginning of the study and not repeated afterward [36].Only few studies showed no variations in the BCVA between the analyzed groups [33,34].

CST Reduction
The studies that included patients without DME generally did not show a significant impact of anti-VEGF therapy as an adjunct to PRP on the CST [27,30,39,41].In the PRIDE study by Lang et al. [41] and subgroup analysis by Shahraki et al. [46], anti-VEGF therapy alone yielded the lowest CST values at the end of follow-up.

Other Data
The other data reported in the analyzed studies showed that PRP yielded more functional retinal damage than did combination therapy or intravitreal therapy alone.Reductions of ERG readings [33,38] as well as deficits in the visual field [46] were greater in the PRP alone group.One study reported improvements in the condition of the ellipsoid zone and external limiting membrane among the PRP plus IVR and IVR monotherapy groups [43].The recent study by Si et al. [48] evaluated changes in angio-OCT vascular parameters after PRP plus and PRP treatments.The foveal avascular zone area and number of microaneurysms decreased more, and the vessel density of the superficial capillary plexus significantly increased in the combination group.In the study by Zhou et al. [36], the times to resorption of VH and regression of NV were significantly shorter with anti-VEGF therapy as an adjunct to PRP (resorption of VH: 12.1 weeks for PRP alone vs. 8.4 for combination treatment, regression of NVD: 15 weeks for PRP alone vs. 12.5 weeks for combination treatment).All these data prove the additional benefits of anti-VEGF injections as an adjunct therapy for PDR.

Conclusions
This review demonstrates that PRP combined with anti-VEGF therapy is superior to PRP alone in the management of PDR.This combination treatment yields better and faster NV regression and a lower incidence of serious complications requiring PPV.Nevertheless, complete NV regression is not achieved with any treatment in 100% of patients.Moreover, further research is needed to establish the most effective schedule for intravitreal injections as an adjunct to PRP.The current literature shows that in some cases, cessation of anti-VEGF injection in combination treatment for PDR can lead to relapse of NV.

Table 1 .
Summary of the findings of the included studies on combination treatment of PDR.