Photodynamic therapy in combination with ranibizumab versus ranibizumab monotherapy for wet age-related macular degeneration: A systematic review and meta-analysis

Accepted Manuscript Title: Photodynamic therapy in combination with ranibizumab versus ranibizumab monotherapy for wet age-related macular degeneration: A systematic review and meta-analysis Authors: Yongxian Su, Jiawei Wu, Yu Gu PII: DOI: Reference:

S1572-1000(17)30538-0 https://doi.org/10.1016/j.pdpdt.2018.05.002 PDPDT 1161

To appear in:

Photodiagnosis and Photodynamic Therapy

Received date: Revised date: Accepted date:

5-12-2017 8-4-2018 4-5-2018

Please cite this article as: Su Y, Wu J, Gu Y, Photodynamic therapy in combination with ranibizumab versus ranibizumab monotherapy for wet age-related macular degeneration: A systematic review and meta-analysis, Photodiagnosis and Photodynamic Therapy (2010), https://doi.org/10.1016/j.pdpdt.2018.05.002 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

IP T

Photodynamic therapy in combination with ranibizumab versus ranibizumab monotherapy for wet age-related macular degeneration: A systematic review and meta-analysis

SC R

Author List Yongxian Sua, Jiawei Wua, Yu Gua a The

Second Affiliated Hospital of Southern Medical University, Guangzhou 510000,

A

N

U

China

Photodynamic therapy combined with ranibizumab has no inferiority compared with monotherapy.



Combination therapy could decrease the injections of ranibizumab, compared with ranibizumab

PT

mono-therapy. 

ED



M

Highlights

Combination therapy had a better stability of vision than monotherapy in one year, although the vision it maintained was not as good as that with monotherapy. The baseline BCVA has a great influence on the number of ranibizumab injections.

CC E

 

BCVA improvement with combination therapy was inferior to that with ranibizumab monotherapy over 12 months of follow-up and its proportions of patients gaining more than 15 letters was

A

smaller than that in monotherapy.

Abstract

Objective To evaluate the efficacy and safety between photodynamic therapy (PDT) combined with intravitreal ranibizumab (IVR) and ranibizumab monotherapy in treating wet age-related macular degeneration (AMD). Methods A systematic search was performed in the PubMed, Embase, Web of Science and the Cochrane

IP T

Library databases through December 31, 2017. The methodological quality of the references was evaluated according to the Cochrane quality assessment. RevMan 5.3 software was used to perform the meta-analysis.

SC R

Results

Eight RCTs involving 817 participants were included. Wet AMD eyes in the mono-group achieved better best-corrected vision acuity (BCVA) than the combination group in month 12 (WMD =-0.19, 95% CI =-0.32 to -0.06, P=0.004, I2=18%). The proportion of patients gaining more than 15 letters from baseline in the mono-group was larger than that in the combination group (RR=0.70, 95% CI:

U

0.56 to 0.87, P=0.001). However, the number of ranibizumab injections with combination therapy was smaller than that with mono-therapy (MD=-1.13, 95% CI: -2.11 to -0.15, P=0.02, I2=85%). No

N

significant differences were observed in the proportions of patients losing more than 15 letters, central retinal thickness (CRT), lesion size of choroidal neovascularization (CNV) and adverse

A

events.

M

Conclusions

Combination therapy decreased the number of injections of ranibizumab, although its BCVA improvement was inferior to that of monotherapy over 12 months of follow-up. Given the inherent

Keywords

PT

findings in this area.

ED

limitations of the included trials, more studies are needed to further validate and update the

Photodynamic therapy; Ranibizumab; Age-related macular degeneration; Systematic review; Meta-

CC E

analysis

A

1. Introduction

Age-related macular degeneration (AMD) is one of the prevailing causes of serious vision loss, mostly

in people older than 55 years old in Western countries [1, 2].Characterized by choroidal neovascularization (CNV)[3], wet AMD covers about 20% of AMD patients, but it is responsible for almost 80%-90% severe vision loss [4-8].The intractability of wet AMD, however, lies in its relationships with age, metabolic changes in the retinal pigment epithelium, oxidative damage, especially light injury, inflammatory and immune disorders, genetic factors and so forth disease in ophthalmology.

[4],

confirming that AMD is an irreversible, chronic, degenerative

Treatments for wet AMD focus on main three targets: neovascularization; angiogenic processes; and inflammatory, cicatricial and exudative processes. The current treatments for neovascularization vary. For example, there are photodynamic therapy, anti-tissue growth factors such as ICON-1, low-dose transpupillary thermotherapy (TPT) and anti-PDGF-β drugs such as pegpleramib. Regarding the angiogenic process, treatments, such as anti-VEGF drugs, steroid agents like dexamethasone, anti-FGF drugs, topical squalamine, tyrosine kinase inhibitors like X-82 and angiopoietin-2 inhibitors, are of great use, and ICON-1, steroid agents and anti-ANG-2 drugs work by reducing exudation. Ranibizumab, an anti-VEGF antibody fragment, can selectively inhibit the biological activity of multiple isoforms of VEGF-A, which is a key factor in the pathogenesis of choroidal neovascularization (CNV),

the pathological vessels effectively after injection because of its small molecular weight

IP T

thus hindering the CNV development process. At the same time, it can easily penetrate the retina to reach [9].

Ranibizumab

is the only agent with sufficient Level I evidence and a Health Canada–approved indication for the

SC R

treatment of wet AMD[10]. Anti-VEGF treatment is often initiated with three loading doses, administered

approximately 1 month apart, or it is administered as needed (pro re nata, PRN)[11].However, the function of the drug is temporary, so ranibizumab is often injected once per month to maintain a steady effect, which can burden patients greatly, as well as have adverse reactions, such as high intraocular pressure, retinal detachment, vitreous haemorrhage, etc [12-14].

U

The mechanism of photodynamic therapy is to cause endothelial cell lesions by the release of free radicals when verteporfin is activated by laser light. These free radicals induce platelet activation and local

N

clot formation, thus achieving vascular occlusion [15, 16]. Remarkable outcomes have been seen in reducing the moderate and severe vision loss in patients with PDT, as some studies have confirmed

A

addition, PDT could be effective when patients are tolerant of anti-VEGF drugs

[17].

[17, 18].

In

However, PDT can

M

also affect the physiology of the choriocapillaris surrounding the pathological CNV lesions and cause hypoxia, thus stimulating the regeneration of VEGF, which anti-VEGF drugs can partly counteract [4]. PDT only works on neovascularization and has little effect on the angiogenic process

[19].

Therefore, PDT can

ED

only maintain the relative stability of vision and can delay macular degeneration lesions, but it cannot significantly improve the visual acuity.

PDT does not work as a single therapy for AMD when feedback and control mechanisms help to

PT

bypass inhibition of a single pathway. Combination approaches in theory allow for the targeting of more than one component of a disease process at a time [20]. Considering PDT and anti-VEGF drugs have different mechanisms of action, it has been hypothesized that combination therapy could have better

CC E

effects and reduce the injections of ranibizumab to lessen the financial burden as well as the side effects caused by the drugs

[17, 19].

Combination therapy consists of photodynamic therapy with verteporfin

performed at month 0 and injections of ranibizumab at month 0, 1, and 2 and then injections of ranibizumab as needed. There exist quite a few studies concerning the effects of this treatment. According to www.amdbook.org, the MONT BLANC study[21] showed no significant differences between the two

A

groups, although the number of treatments was slightly smaller with the combined treatment. Similarly, in the DENALI study, the overall benefit for patients was a reduced frequency of Lucentis over at least three months during the study[22]. Moreover, there have been no definite conclusions about whether ranibizumab monotherapy is superior to PDT combined with ranibizumab in treating nAMD. Moreover, whether combination therapy can reduce the number of ranibizumab injections remained unknown in the two latest meta-analysis

[23, 24].

In Si J. K.’s study

[23]

due to inadequate data on ranibizumab retreatment, analysis

of the numbers of injections could not be performed. Tong Y’s study

[24]showed

that there was no

significant difference between the groups in numbers of ranibizumab treatments at month 12. Considering

all of above evidence, we conducted a new systematic review and meta-analysis to evaluate the efficacy and safety of combination therapy compared to ranibizumab monotherapy.

2. Methods 2.1 Search strategy A systematic search was performed in the PubMed, Embase, Web of Science and the Cochrane Library databases through December 31, 2017. We adopted exploded Medical Subject Headings (MeSH) terms and corresponding keywords in the electronic search process. The search terms were the following :(MeSH exp Macular Degeneration, and keywords), and (MeSH exp Ranibizumab, and keywords Lucentis) and (MeSH exp Photochemotherapy, and keywords), and (Publication Type exp Randomized

IP T

Controlled Trial, and keywords). No language or publication year restrictions were imposed. We also manually checked the entire references of the selected studies to identify potentially eligible studies. Conference summaries were also searched to avoid omitting relevant studies.

SC R

2.2 Inclusion and exclusion criteria

The following criteria were used to select the qualified studies: 1) RCTs compared ranibizumab and PDT versus ranibizumab alone in treating wet AMD; 2) all of the patients were diagnosed with wet AMD by professional ophthalmic examinations; 3) CRT assessed by optical coherence tomography (OCT) and BCVA assessed by ETDRS charts; 4) for studies published by the same group regarding the same

U

population, only the most recent reports or the reports with the largest sample size were included; and 5) BCVA, CRT, and number of ranibizumab injections should be included in one RCT.

N

The exclusion criteria were as follows: 1) retrospective, preclinical studies, reviews, and meta-analysis not containing original data; 2) patients with previous intravitreal treatment, drug treatment or laser

M

other than age-related macular degeneration.

A

treatment in the study eye within 30 days before enrolment; and 3) studies of CNV secondary to causes

2.3 Data exaction and quality assessment

Three authors independently used the aforementioned inclusion/exclusion criteria to select studies and

ED

then extract data. Data from the qualified studies included: 1) clinical characteristics of patients, such as age and numbers; 2) information about the studies, such as publication year, center location, treatment protocol, inclusion and exclusion criteria and re-treatment criteria; 3) primary outcomes: BCVA, number

PT

of ranibizumab injections and CRT; and secondary outcomes: lesion size of CNV, proportion of patients gaining ≥ 15 letters, proportion of patients losing ≥ 15 letters and ocular adverse events. If these data could not be directly extracted from the study, much work was done to obtain them, including sending e-

CC E

mail to the author, researching the associated conference summaries and other studies citing the study in question or searching for the data at clinicaltrials.gov. Risk of bias was assessed with the risk-of bias tool provided by Cochrane quality assessment. 2.4 Statistical analysis This meta-analysis was performed by RevMan 5.3 (The Cochrane Collaboration, London, UK). We

A

calculated the mean differences (MDs) and 95% confidence intervals (95% CIs) for continuous variables, as well as risk ratios (RRs) and 95% confidence intervals (CIs) for the dichotomous outcomes. If the data were shown in different form, we used the weighted mean difference (WMD) with 95% confidence intervals (95% CIs) for continuous variables. The clinical heterogeneity was evaluated depending on the baseline characteristics and the statistical heterogeneity was assessed by I-square (I2) statistic. If heterogeneity was small (I2< 50%), the fixed effects model was used; otherwise, the random effects model was used.

3. Results 3.1 Identification and selection of studies

In total, 198 articles were identified initially; 57 duplicates and 81 unrelated articles were excluded after screening titles and abstracts. The remaining articles underwent detailed evaluations. After an independent review, 52 articles were excluded for several reasons. Finally, 8 RCTs were included

[25-32].

PT

ED

M

A

N

U

SC R

IP T

The reference search process and exclusion reasons are summarized in Figure 1.

Table 1. Flow diagram of selection of RCTs

CC E

Abbreviations: RCT: randomized controlled trial; PDT: photodynamic therapy; AMD: age-related macular degeneration

3.2 Study characteristics and quality assessment A total of 922 participants (932 eyes) were included in the eight studies. Hatz’s study[25] included 19

patient in combination therapy and 20 in monotherapy, but there remained 18 in combination therapy and 19 in monotherapy at the end because of adverse events and unwillingness to continue. Krebs’s study

A

included 24 patients in combination therapy and 24 in monotherapy, but there remained 19 patients and 22 patients finally because of withdrawals. Larsen’s study [29] included 19 patients in combination therapy and 20 in monotherapy, but there remained 18 patients and 19 patients finally. Kaiser’s study[30] was separated into three groups, namely verteporfin standard fluence (SF) plus ranibizumab, verteporfin reduced fluence (RF) plus ranibizumab and ranibizumab. Semeraro’s study [27] was separated into three groups, namely verteporfin reduced fluence (RF) plus ranibizumab, ketorolac eye drops plus ranibizumab and ranibizumab monotherapy. Bashshur’s study[31] included 30 patients but 40 eyes in total. The requirements regarding BCVA in study eyes varied. The BCVA in study eyes for these five RCTs[25, 26, 29, 30, 32]

was between 73 and 24 letters (20/40 to 20/320 Snellen equivalent), while it was from 20/50 to

20/200 Snellen equivalent in Bashshur’s study[31] and more than 33 letters in Krebs’s study[28]. However, Semeraro’s study[27] had no requirement regarding BCVA in the study eyes enrolled. Moreover, the dose of PDT was low (25J/cm2) in Semeraro’s study and Kaiser’s study established two doses of PDT (25J/cm2 and 50J/cm2), while it was 50J/cm2 in the other studies. There was no sham PDT in the ranibizumab monotherapy group in three RCTs quality of these five

[26, 28, 31],

RCTs[25, 28-30, 32]

and the blinding of Semeraro’s study was open label. The

was the highest, and the quality of Semeraro’ study was the lowest.

The quality of Bashshur’s study and Weingessel’s study[26] was moderate. The characteristics of all of the RCTs are summarized in Table 1, and quality assessment results (Figure 2) of all of the included studies

N

U

SC R

IP T

are presented below.

A

Figure 2. Risk of bias graph. Notes: (A) Each risk of bias item for each included study. (B) The overall risk

CC E

PT

ED

M

of bias is relatively low. “+” indicates yes; “?” indicates not clear

A

3.3 Best corrected visual acuity

Eight studies were included in this section (Figure 3). No heterogeneity in BCVA was observed among

the studies at month 3 and month 12, but at month 6, the heterogeneity was obvious (P=0.06, I2=65%). No significant difference was observed between combination therapy and monotherapy at month 3 and 6 (WMD =−0.04, 95% CI: −0.22 to 0.13, P=0.62, I2 =0%; WMD =-0.15, 95% CI: -0.52 to 0.22, P=0.43, I2=65%). However, there was a significant difference with combination therapy at month 12 (WMD =-0.19, 95% CI =-0.32 to -0.06, P=0.004, I2=18%). This result suggested that ranibizumab monotherapy achieved

better BCVA improvement than combination therapy as an AMD treatment. 3.4 Proportion of patients gaining ≥ 15 letters Six hundred fifty-six patients (661 eyes) were included in this section (Figure 5). No heterogeneity was observed (P=0.91, I2 =0%). The analysis showed a significant difference between the two groups in the proportion of patients gaining ≥ 15 letters (RR=0.70, 95% CI: 0.56 to 0.87, P=0.001), showing that the proportion of patients gaining ≥ 15 letters in combination therapy was statistically smaller than those in the mono-therapy group after 12 months. 3.5 Proportion of patients losing ≥ 15 letters

IP T

Seven hundred seven eyes (697 patients) were included in this section (Figure 6). No heterogeneity

was observed among the studies (P=0.65, I2 =0%). The analysis showed no significant difference between

SC R

the two groups in proportion of patients losing ≥ 15 letters (RR=1.35, 95% CI: 0.89 to 2.04, P=0.16).

N

U

A

M

A

B

CC E

PT

ED

C

Figure 3. Forest plot of standard mean difference in BCVA (logMAR was used in Semeraro’s study and ETDRS letters in the others). A: BCVA in month 3; B: BCVA in month 6; C: BCVA in month 12.

A

3.6 Central retinal thickness Mean CRT change was assessed in all of the studies (Figure 7). No heterogeneity was observed among

the studies (P=0.31, I2=15%). The analysis showed no significant difference between the two groups in CRT (MD =4.80, 95% CI: -6.28 to 15.89, P=0.40). 3.7 Choroidal neovascularization Three studies were included in this section (Figure 9). No heterogeneity was observed among the studies (P=0.72, I2 =0%). The analysis showed no significant difference between the two groups in lesion size of CNV (MD =0.16, 95% CI: -0.56 to 0.88, P=0.66). 3.8 Number of ranibizumab rejections

The analysis showed a significant difference between the two groups in the number of ranibizumab injections (MD=-1.13, 95% CI: -2.11 to -0.15, P=0.0002, I2=85%). Subgroup analysis was conducted according to BCVA baseline. In subgroup 1 within the limitation of 73-24 letters, the heterogeneity declined (P=0.70, I2=0%) and there was no significant difference in the number of ranibizumab injections (WMD: 0.35; 95% CI: -0.78 to 0.09; P=0.12). In subgroup 2 without serious limitation in BCVA baseline, the heterogeneity declined (P=0.89, I2=0%), and there was a significant difference in the number of ranibizumab injections (WMD: -1.98; 95% CI: -2.56 to -1.40; P=0.00001). 3.9 Adverse events No heterogeneity was observed among the four studies (P=0.30, I2 =18%). The analysis showed no

SC R

IP T

significant difference between the two groups in adverse events (RR=1.12, 95% CI: 0.94 to 1.33, P=0.22).

ED

M

A

N

U

Figure 4. Forest plot of standard mean difference in BCVA of two studies with RF PDT at month 12

A

CC E

PT

Figure 5. Risk ratio of proportion of patients gaining ≥ 15 letters at month 12

Figure 6. Risk ratio of proportion of patients losing ≥ 15 letters at month 12

3.10 Sensitivity analysis and publication bias Sensitivity analysis was performed to assess the influence of each study by the sequential removal of each one. No heterogeneity was observed in addition to the BCVA subgroup at month 6 and the number of ranibizumab injections at month 12. The heterogeneity of the BCVA in month 6 declined (P=0.32, I 2 =0%) after Hatz’s study was excluded, probably because the dosage of ranibizumab in this study was 0.3mg while in the others, it was 0.5mg. Regarding ranibizumab injections, sensitivity analysis revealed

that the changes in P were not obvious by removing each study, respectively. No funnel plot could be

U

SC R

Figure 7. Forest plot of mean difference of CRT at month 3, 6, and 12

IP T

drawn for the meta-analysis because there were only eight studies.

ED

M

A

N

Figure 8. Forest plot of mean difference in CRT of two studies with RF PDT at month 12

A

CC E

PT

Figure 9. Forest plot of mean difference in lesions of CNV at month 12

Figure 10. Forest plot of number of ranibizumab injections at month 12

Figure 11. Risk ratio of adverse events over 12 months

4. Discussions We found that combination therapy showed inferiority to ranibizumab mono-therapy in mean BCVA change from baseline at month 12 (WMD =-0.19, 95% CI =-0.32 to -0.06, P=0.004, I2=18%), which was the opposite of the outcomes that we expected. We found that, in these two studies

[26, 28],

the figure of

mean BCVA change from baseline was negative in the combination group while it was positive in the mono-group. Further, in most studies, both the combination group and the mono-group had better mean BCVA improvement from baseline, but the improvement in the combination therapy group was slighter better than that in the mono-therapy group (Figure 2), perhaps because of the VEGF rebound effect caused by PDT when blocking new vessels. Exactly because of the obstruction of CNV vessels and low

IP T

perfusion by PDT in the treatment area, inflammatory mediators and free radicals were released, resulting

in new reproduction of pathological vessels. Further, ranibizumab can only inhibit VEGF; therefore, it was unavoidable that the combination group might have weakened the effect of ranibizumab. We also found

SC R

differences between these two studies that PDT was conducted after ranibizumab, in contrast to the other 6 studies. Debefve E.’s study[33]concerning the optimal time sequence suggested that PDT was supposed to be added 1 day before ranibizumab, which could achieve better vision and fewer ranibizumab injections. It was assumed that ranibizumab added later could reverse the defect to a greater extent, although not entirely.

U

To determine whether reduced fluence of PDT might work given the premise that PDT was conducted before ranibizumab, we conducted two analyses containing studies with 50% reduced PDT(25J/cm2),

N

considering that the threshold fluence dose of vPDT is approximately 20% of standard fluence

[7]

(Figure

4, Figure 8). No difference could be seen between the two groups in BCVA and CRT (MD =-0.19, 95% CI:

A

-0.43 to 0.053, P=0.12; MD =14.38, 95% CI: -14.58 to 43.33, P=0.33). To solve this question, more clinical

M

evidence is needed. There was significant heterogeneity in mean BCVA changes at month 6 (I2=65%). However, subgroup analysis could not be conducted with only three RCTs. The former meta-analysis

[23]

found that the combination group tended to have a greater reduction in

ED

CRT (MD= -4.13μm; 95%CI, -25.88 to 17.63, =0.71), but our meta-analysis showed no reduction tendency of CRT in the combination group (MD =4.80, 95% CI: -6.28 to 15.89, P=0.40). Additionally, there was a significant difference in mean BCVA improvement between the groups, but no difference in CRT. The

PT

reason for this outcome might be the following. First, the number of included RCTs might matter. Secondly, it might have been the different OCTs in the 8 studies (Table1). Last but not least, there was no absolute correlation between the objective index, CRT and the subjective index, BCVA. Ou WC’s study [34] showed

CC E

that the relationship between BCVA and CRT was more complex in AMD and might involve factors such as atrophy and bleeding. All of these factors indicated that combination therapy was inferior to monotherapy in terms of vision

improvement. However, it was recognized that polypoidal choroidal vasculopathy (PCV) might constitute as much as 50% of cases of wet AMD in some Asian countries

[19],

and some Asian patients with PDT

A

have had improved vision and less frequent repeat treatments, compared with white patients, likely because of more retinal pigment and higher PDT sensitivity in darker-skinned people. Further, a metaanalysis

[35]

showed that ranibizumab combined with PDT in PCV patients was superior to ranibizumab

monotherapy for vision, indicating the sensitivity of PDT. However, it remains a controversy whether PCV is a subtype of nAMD; in addition, in the processing of searching the literature, we could not find the trials involving Asian patients under our inclusion and exclusion criteria as well. To make the findings more credible and rigorous, more eligible trials are needed to improve this analysis. Importantly, our meta-analysis found that combination therapy could effectively reduce the number of

ranibizumab injections versus monotherapy (MD=-1.13, 95% CI: -2.11 to -0.15, P=0.02, I2=85%), compared with former meta-analyses

[23, 24].

Subgroup analysis was conducted according to the BCVA

baseline. The BCVA baselines in Larsen’s study and Weingessel’s study were between 73 and 24 letters (20/40 to 20/320 Snellen equivalent), while it was more than 33 letters in Krebs’s study and Semeraro’s study had no requirement on BCVA in the study eye enrolled. This outcome illustrated that the baseline BCVA had a great influence on the number of ranibizumab injections. Hernandez[36] noted that ranibizumab, when used for two consecutive years, cost at least EUR 131,275 with a higher surcharge, while PDT cost up to EUR 30,000. At the same time, as the number of ranibizumab injections decreased, the incidence of side effects decreased correspondingly. Furthermore, there can be no significant

IP T

improvement in BCVA in some patients with wet AMD receiving monotherapy [37], possibly because of the

small proportion of new angiogenesis factors in the pathogenesis of AMD or tolerance after long-term use of anti-angiogenic drugs. It is suggested that the use of combination treatment regimens could be a

who are tolerant to ranibizumab

SC R

secondary option for patients who are not able to be followed up on time, who have financial burdens or [38].

Moreover, retreatment with ranibizumab should be increased to maintain the vision since the mean BCVA and CRT improvement in the combination therapy were inferior to those with monotherapy. However, the number in the combination group was reduced by 1.13 compared with the mono-group statistically.

U

We noted that the vision was worse than baseline in the combination group in some studies (Figure 7) with the limitation of retreatments that BCVA lost should be more than 5 letters and CRT increased

N

by >100μm. That is, when the vision decreased, but the decreased vision did not meet to the criteria, no treatment was performed. It can also be seen that mean BCVA and CRT improvement did exist in the

A

combination group although they were slighter. Therefore, we can conclude that combination therapy

M

resulted in better stability of vision than mono-group at one year, although the vision that it maintained was not as good as that of the mono-group. Certainly, some clinical studies have confirmed tolerance of ranibizumab. Although the early use of anti-VEGF drugs alone can reduce subretinal fluid and prevent

ED

early CNV progression, the effect of this drug on the development of late CNV greatly diminishes after angiogenesis[39-43]. Ozkaya A.’s study[18] showed that half of the patients with nAMD treated with ranibizumab monotherapy could not maintain their visual acuity after 5 years of follow-up. Since the

PT

longest time point of our analysis was one year, it was not sufficient to confirm the tolerance of ranibizumab or whether ranibizumab monotherapy still achieved better BCVA than combination therapy over longer follow-ups. More trials with longer follow-ups are expected.

CC E

There was also no difference in adverse ocular events between the groups (RR=1.2, 95% CI: 0.94 to 1.33, P=0.22, I2=18%). Four RCTs [25, 27, 29, 30] reported the proportions of adverse reactions in each group, three of which[27, 29, 30] were classified in detail. The most common ocular adverse events in these 3 RCTs were ocular pain and retinal haemorrhage, followed by conjunctivitis, eyeball congestion and myodesopsia. Larsen’s study also reported adverse extraocular events including hypertension and nasopharyngitis.

A

Vallance’s study showed no adverse reactions at the end of the research, and Bashshur’s study only described the adverse reactions but did not report the proportions. There was no difference in the proportions of adverse reactions between the two groups. Further scientific trials are needed to solve this problem. Combined therapies have been popular for years. The Lucedex study[44] combined PDT with ranibizumab and dexamethasone, and the results showed that the triple therapy significantly reduced the lesion size of CNV, compared to ranibizumab monotherapy. The EMERGE study showed that ICON-1 combined with ranibizumab could not only eliminate CNV but could also reduce exudation. Similarly, anti-

VEGF drugs combined with anti-FGF drugs could inactivate these growth factors and avoid endothelial cell sprouting

[22].

Soderberg’s study[45] showed that low-dose TPT combined with ranibizumab could

reduce the number of ranibizumab injections in 2 years. The tyrosine kinase inhibitor X-82, which could inhibit all receptor subtypes of VEGF and PDGF, could achieve low doses and low risks when combined with ranibizumab[46]. Wroblewski’s study

[47]

showed that ranibizumab combined with topical squalamine

could achieve better visual function than ranibizumab monotherapy, but there were also some combination therapies that were not satisfatory. For example, aflibercept coformulated with rinucumab in Capella’s study, regardless of BCVA, CRT or adverse effects, was inferior to aflibercept monotherapy.Additionally, there was no difference between pegpleranib combined with Eylea, Eylea monotherapy and pegpleranib

IP T

monotherapy for 12 months in the Fovista Trial. Currently, the most promising strategy is angiopoietin pathway inhibition, in which inhibition of ANG-2 or activation of Tie2 is worth concentration.

Our meta-analysis had the following strengths: 1) our meta-analysis found that combination therapy

SC R

could effectively reduce the number of ranibizumab injections compared with the mono-group (MD=-1.13,

95% CI: -2.11 to -0.15, P=0.02, I2=85%); 2) our inclusion and exclusion criteria were stricter than the those of former meta-analysis; and 3) we have discussed the optimal time sequence of photodynamic therapy and ranibizumab. Our meta-analysis also had the following limitations: 1) most studies failed to mention the method of allocation concealment, so the quality of these studies was moderate; 2) some studies did

U

not mention the proportion of each type of CNV, and PDT was more suitable for the classical type, while anti-VEGF drugs were fit for all types; 3) no funnel plots could be drawn for the meta-analysis because

N

there were only eight studies; 4) because there were only three RCTs measuring mean BCVA changes at month 6, we were not able to perform the subgroup analysis; 5) it would be better to include data for more

A

years because wet AMD is a chronic disease, and therefore a long-term perspective is needed; and 6)

M

the types of OCT differed in the 8 RCTs, so there might be statistical errors in the CRT data. The differences between this meta-analysis and the former two are shown in Table 2.

5. Conclusion

ED

Although BCVA improvement in the combination group was inferior to that with ranibizumab alone at month 12 and the proportion of patients gaining more than 15 letters was less than that of the mono-group, PDT combined with ranibizumab could decrease the number of injections of ranibizumab, thus reducing

PT

the financial burden and making it more convenient for patients who could not be regularly followed up. We should consider individualized treatments according to patients’ specific conditions and different needs. There was no difference in adverse effects between the groups. More clinical studies are needed to verify

CC E

the safety of both treatments.

A

References

1.

Kawczyk-Krupka, A., et al., Vascular-targeted photodynamic therapy in the treatment of neovascular age-related macular degeneration: Clinical perspectives. Photodiagnosis Photodyn Ther, 2015. 12(2): p. 161-75.

2.

Li, S., et al., Combinatorial treatment with topical NSAIDs and anti-VEGF for age-related macular degeneration, a meta-analysis. PLoS One, 2017. 12(10): p. e0184998.

3.

Sivaprasad, S. and P. Hykin, What is new in the management of wet age-related macular

degeneration? Br Med Bull, 2013. 105: p. 201-11. 4.

Zarbin, M.A., Current concepts in the pathogenesis of age-related macular degeneration. Arch Ophthalmol, 2004. 122(4): p. 598-614.

5.

Tuuminen, R., A. Tuulonen, and K. Kaarniranta, The Finnish national guideline for diagnosis, treatment and follow-up of patients with wet age-related macular degeneration. Acta Ophthalmol, 2017. 95(7): p. 649-650.

6.

Colquitt, J.L., et al., Ranibizumab and pegaptanib for the treatment of age-related macular degeneration: a systematic review and economic evaluation. Health Technol Assess, 2008. 12(16): p. iii-iv, ix-201. Chen, E., et al., Lucentis using Visudyne study: determining the threshold-dose fluence of

IP T

7.

verteporfin photodynamic therapy combined with intravitreal ranibizumab for exudative macular degeneration. Clin Ophthalmol, 2010. 4: p. 1073-9.

Dong, Y., et al., Effect of anti-VEGF drugs combined with photodynamic therapy in the

SC R

8.

treatment of age-related macular degeneration. Exp Ther Med, 2016. 12(6): p. 3923-3926. 9.

Michels, S. and P.J. Rosenfeld, [Treatment of neovascular age-related macular degeneration with Ranibizumab/Lucentis]. Klin Monbl Augenheilkd, 2005. 222(6): p. 480-4.

10.

Cruess, A., et al. Canadian expert consensus: Optimal treatment of neovascular age-related

U

macular degeneration. Canadian Journal of Ophthalmology, 2012. 47, 227-35 DOI: 10.1016/j.jcjo.2012.03.007.

Tuuminen, R., et al., The Finnish national guideline for diagnosis, treatment and follow-up of

N

11.

patients with wet age-related macular degeneration. Acta Ophthalmol, 2017. 95(A105 Hodge, W., et al., Pharmacologic management of neovascular age-related macular

M

12.

A

Suppl): p. 1-9.

degeneration: systematic review of economic evidence and primary economic evaluation. Can J Ophthalmol, 2010. 45(3): p. 223-30.

Rosenfeld, P.J., et al., Ranibizumab for neovascular age-related macular degeneration. N Engl

ED

13.

J Med, 2006. 355(14): p. 1419-31. 14.

Brown, D.M., et al., Ranibizumab versus verteporfin for neovascular age-related macular

PT

degeneration. N Engl J Med, 2006. 355(14): p. 1432-44. 15.

Bressler, N.M., Photodynamic therapy of subfoveal choroidal neovascularization in agerelated macular degeneration with verteporfin: two-year results of 2 randomized clinical

CC E

trials-tap report 2. Arch Ophthalmol, 2001. 119(2): p. 198-207. 16.

Augustin, A.J., S. Puls, and I. Offermann, Triple therapy for choroidal neovascularization due to age-related macular degeneration: verteporfin PDT, bevacizumab, and dexamethasone. Retina, 2007. 27(2): p. 133-40.

A

17. 18. 19.

Shah, G.K., D.N. Sang, and M.S. Hughes, Verteporfin combination regimens in the treatment of neovascular age-related macular degeneration. Retina, 2009. 29(2): p. 133-48. Ozkaya, A., et al., Five-year Outcomes of Ranibizumab in Neovascular Age-related Macular Degeneration: Real Life Clinical Experience. Korean J Ophthalmol, 2017. 31(5): p. 424-430. Koh, A., et al., Optimising the management of choroidal neovascularisation in Asian patients: consensus on treatment recommendations for anti-VEGF therapy. Singapore Med J, 2011. 52(4): p. 232-40.

20.

Spaide, R.F., Rationale for combination therapy in age-related macular degeneration. Retina, 2009. 29(6 Suppl): p. S5-7.

21.

Larsen, M., et al., Verteporfin plus ranibizumab for choroidal neovascularization in agerelated macular degeneration: twelve-month MONT BLANC study results. Ophthalmology, 2012. 119(5): p. 992-1000.

22.

Stahl, A., et al., Combinatory inhibition of VEGF and FGF2 is superior to solitary VEGF inhibition in an in vitro model of RPE-induced angiogenesis. Graefes Arch Clin Exp Ophthalmol, 2009. 247(6): p. 767-73.

23.

Si, J.K., et al., Combination of ranibizumab with photodynamic therapy vs ranibizumab monotherapy in the treatment of age-related macular degeneration: a systematic review and meta-analysis of randomized controlled trials. Int J Ophthalmol, 2014. 7(3): p. 541-9. Tong, Y., et al., Comparison of the efficacy of anti-VEGF monotherapy versus PDT and

IP T

24.

intravitreal anti-VEGF combination treatment in AMD: a Meta-analysis and systematic review. Int J Ophthalmol, 2016. 9(7): p. 1028-37.

Hatz, K., et al., Ranibizumab plus Verteporfin Photodynamic Therapy in Neovascular Age-

SC R

25.

Related Macular Degeneration: 12 Months of Retreatment and Vision Outcomes from a Randomized Study. Ophthalmologica, 2015. 233(2): p. 66-73. 26.

Weingessel, B., K. Mihaltz, and P.V. Vecsei-Marlovits, Predictors of 1-year visual outcome in OCT analysis comparing ranibizumab monotherapy versus combination therapy with PDT in

U

exsudative age-related macular degeneration. Wiener Klinische Wochenschrift, 2016. 128(15-16): p. 560-565.

Semeraro, F., et al., Treatment of Exudative Age-Related Macular Degeneration with

N

27.

Ranibizumab Combined with Ketorolac Eyedrops or Photodynamic Therapy. Retina-the Krebs, I., et al., Comparison of Ranibizumab monotherapy versus combination of Ranibizumab

M

28.

A

Journal of Retinal and Vitreous Diseases, 2015. 35(8): p. 1547-1554. with photodynamic therapy with neovascular age-related macular degeneration. Acta Ophthalmologica, 2013. 91(3): p. E178-E183. Larsen, M., et al., Verteporfin plus Ranibizumab for Choroidal Neovascularization in Age-

ED

29.

related Macular Degeneration Twelve-month MONT BLANC Study Results. Ophthalmology, 2012. 119(5): p. 992-1000.

Kaiser, P.K., et al., Verteporfin plus Ranibizumab for Choroidal Neovascularization in Age-

PT

30.

Related Macular Degeneration Twelve-Month Results of the DENALI Study. Ophthalmology, 2012. 119(5): p. 1001-1010. Bashshur, Z.F., et al., Ranibizumab monotherapy versus single-session verteporfin

CC E

31.

photodynamic therapy combined with as-needed ranibizumab treatment for the management of neovascular age-related macular degeneration. Retina, 2011. 31(4): p. 63644.

A

32.

33.

Vallance, J.H., et al., A randomised prospective double-masked exploratory study comparing combination photodynamic treatment and intravitreal ranibizumab vs intravitreal ranibizumab monotherapy in the treatment of neovascular age-related macular degeneration. Eye, 2010. 24(10): p. 1561-1567. Debefve, E., et al., Combination therapy using verteporfin and ranibizumab; optimizing the timing in the CAM model. Photochem Photobiol, 2009. 85(6): p. 1400-8.

34.

Ou, W.C., et al., Relationship Between Visual Acuity and Retinal Thickness During AntiVascular Endothelial Growth Factor Therapy for Retinal Diseases. Am J Ophthalmol, 2017. 180: p. 8-17.

35.

Liu, L., et al., Photodynamic therapy in combination with Ranibizumab versus Ranibizumab monotherapy for polypoidal choroidal vasculopathy: A systematic review and meta-analysis. Photodiagnosis Photodyn Ther, 2017.

36.

Hernandez-Pastor, L.J., et al., Cost-effectiveness of ranibizumab compared with photodynamic treatment of neovascular age-related macular degeneration. Clin Ther, 2008. 30(12): p. 243651.

37.

Teper, S.J., et al., Photodynamic therapy in VEGF inhibition non-responders-Pharmacogenetic study in age-related macular degeneration assessed with swept-source optical coherence tomography. Photodiagnosis Photodyn Ther, 2016. 13: p. 108-13. Stewart, M.W., Individualized Treatment of Neovascular Age-Related Macular Degeneration: What are Patients Gaining? Or Losing? J Clin Med, 2015. 4(5): p. 1079-101.

39.

Bradley, J., M. Ju, and G.S. Robinson, Combination therapy for the treatment of ocular

SC R

neovascularization. Angiogenesis, 2007. 10(2): p. 141-8. 40.

IP T

38.

Bergers, G., et al., Benefits of targeting both pericytes and endothelial cells in the tumor vasculature with kinase inhibitors. J Clin Invest, 2003. 111(9): p. 1287-95.

41.

Peters, S., et al., Antipermeability and antiproliferative effects of standard and frozen bevacizumab on choroidal endothelial cells. Br J Ophthalmol, 2007. 91(6): p. 827-31. Volcker, M., et al., [Early antiexudative response--OCT monitoring after intravitreal

U

42.

bevacizumab injection]. Ophthalmologe, 2006. 103(6): p. 476-83.

Kaiser, P.K., et al., Angiographic and optical coherence tomographic results of the MARINA

N

43.

study of ranibizumab in neovascular age-related macular degeneration. Ophthalmology, Ranchod, T.M., et al., LuceDex: a prospective study comparing ranibizumab plus

M

44.

A

2007. 114(10): p. 1868-75.

dexamethasone combination therapy versus ranibizumab monotherapy for neovascular agerelated macular degeneration. Retina, 2013. 33(8): p. 1600-4. Soderberg, A.C., et al., Combination therapy with low-dose transpupillary thermotherapy and

ED

45.

intravitreal ranibizumab for neovascular age-related macular degeneration: a 24-month prospective randomised clinical study. Br J Ophthalmol, 2012. 96(5): p. 714-8. Diago, T., et al., Ranibizumab combined with low-dose sorafenib for exudative age-related

PT

46.

macular degeneration. Mayo Clin Proc, 2008. 83(2): p. 231-4. 47.

Wroblewski, J.J. and A.Y. Hu, Topical Squalamine 0.2% and Intravitreal Ranibizumab 0.5 mg as

CC E

Combination Therapy for Macular Edema Due to Branch and Central Retinal Vein Occlusion: An Open-Label, Randomized Study. Ophthalmic Surg Lasers Imaging Retina, 2016. 47(10): p.

A

914-923.

Table 1. Characteristics of included studies

Autho Treatment protocol r

Retreatment criteria

Combination therapy

Monothearpy

IP T

Weing 1)0.5mg IVR at month 0,1,2 1)0.5mg IVR at 1)BCVA worsen<5 letters; essel 2)SF verteporfin PDT 1 day following at month 0,1,2 CRT increased>100μm month 0 2)IVR PRN from 2)new macular 3)IVR PRN from month 3 to 11 month 3 to 11 hemorrhage 3)new intra- or subretinal fluid

1)0.5mg IVR at month 0,1,2 1)0.3mg IVR at 1)BCVA worsen<5 letters; 2)SF verteporfin PDT 1 hour before IVR month 0,1,2 CRT increased>100μm 0.3mg at month 0 2)IVR PRN from 2)new intra- or subretinal 3)SF verteporfin PDT the same day at month 3 to 11 fluid month 0

ED

M

Hatz

A

N

U

SC R

Semer A:1)0.5mg IVR at month 0,1,2 1)0.5mg IVR at 1)new macular aro 2)RF verteporfin PDT 1 hour month 0,1,2 hemorrhage before IVR 0.5mg at month 0 2)IVR PRN from 2)new intra- or subretinal 3)IVR PRN from month 3 to 11 month 3 to 11 fluid B:1)0.5mg IVR at month 0,1,2 3)expansion of CNV 2)IVR PRN from month 3 to 11 4)worsen BCVA 3)ketorolac I drop of 3 times a day over 12 months

CC E

PT

Krebs 1)0.5mg IVR at month 0,1,2 1)0.5mg IVR at 1)BCVA worsen<5 letters; 2)SF verteporfin PDT 1 day following at month 0,1,2 CRT increased>100μm month 0 2)IVR PRN from 2)new macular 3)IVR PRN from month 3 to 11 month 3 to 11 hemorrhage 3)new intra- or subretinal fluid

A

Larsen 1)0.5mg IVR at month 0,1,2 1)0.5mg IVR at 1)BCVA worsen<5 letters; 2)SF verteporfin PDT 1 hour before IVR month 0,1,2 CRT increased>100μm 0.5mg at month 0 2)IVR PRN from 2)new macular 3)IVR and PDT PRN from month 3 to 11 month 3 to 11 hemorrhage at intervals of ≥30 and 90 days 3)new intra- or subretinal fluid 4)leakage of FA

IP T

Kaiser A:1)0.5mg IVR at month 0,1,2 1)0.5mg IVR at 1)BCVA worsen<5 letters; 2)SF verteporfin PDT 1 hour month 0-11 CRT increased>100μm before IVR 0.5mg at month 0 2)new macular 3)IVR and PDT PRN from month 3 hemorrhage to 11 at intervals of ≥30 and 90 days 3)cystaid macular edema B:1)0.5mg IVR at month 0,1,2 4)increased PED>100μm 2)RF verteporfin PDT 1 hour before IVR 0.5mg at month 0 3)IVR and PDT PRN from month 3 to 11 at intervals of ≥30 and 90 days

U

SC R

Bashs 1)0.5mg IVR at month 0,1,2 1)0.5mg IVR at 1)BCVA worsen<5 letters; hur 2)SF verteporfin PDT 1 hour before IVR month 0,1,2 CRT increased>100μm 0.5mg at month 0 2)IVR PRN from 2)new macular 3)IVR PRN from month 3 to 11 month 3 to 11 hemorrhage 3)new intra- or subretinal fluid 4)expansion of CNV

Abbreviation:

PT

ED

M

A

N

Vallan 1)0.5mg IVR at month 0,1,2 1)0.5mg IVR at 1)BCVA worsen<5 letters; ce 2)SF verteporfin PDT the same day at month 0,1,2 CRT increased>100μm month 0 2)IVR PRN from 3)IVR and PDT PRN from month 3 to 11 month 3 to 11 at intervals of ≥30 and 90 days

IR: intravitreal ranibizumab; SF: standard fluence; RF: reduced fluence; BCVA: best-corrected vision acuity; CRT: central retinal thickness;

A

CC E

CNV: choroidal neovascularization; FA: fluorescein angiography; PRN: pro re nata

I N U SC R

This meta

The number of

8

studies

Jun-Kang Si 2014

8

7

1) All patients had a professional ophthalmic

1) Included human eyes with active

versus ranibizumab alone in treating AMD;

examination and were diagnosed as AMD;

CNV secondary to AMD;

2) All patients had a professional ophthalmic

2) The study design was limited to RCT;

2) RCTs which compared combination

examination and were diagnosed as AMD;

3) Interventions included anti-VEGF monotherapy

of ranibizumab with PDT ranibizumab

3) The central retinal thickness (CRT) assessed

(inner ocular injection with ranibizumab or

monotherapy;

by optical coherence tomography (OCT) and

bevacizumab) and combined PDT and anti-VEGF

3) Studies reported one or more of the

best corrected vision acuity (BCVA) assessed by

therapy, and the time of follow-up was at least 12

following outcomes: best-corrected

ETDRS charts;

month;

visual acuity (BCVA), central retinal

4) For studies published by the same group

4) Endpoints included at least one of the following: the

thickness (CRT), number of treatments,

regarding the same population, only the most

best-corrected visual acuity (BCVA), CRT, number of

and ocular or systemic adverse events.

recent report or the report with the largest

treatments and proportion of patients who gained 15,

sample size was included for the analysis.

10, 5, or 0 letters of BCVA at 12th month;

PT

CC E A

Yang Tong 2016

1) RCTs comparing ranibizumab and PDT

ED

Major inclusion criteria

hs

M

Differences

A

Table 2. The differences between this meta-analysis and the former meta- analyses.

5) Raw data were available;

Major exclusion

1) Patients with previous intravitreal treatment,

criteria

drug treatment and laser treatment in the study

controlled trials;

eye within 30 days before enrolment;

2) Studies of CNV secondary to causes

2) Lack of complete data of BCVA and CRT;

other than AMD.

3) Studies of CNV secondary to causes other

Not available

1) Studies which were not randomized

I The type of

Ranibizumab only

antibody Main

BCVA

CC E

polyps BCVA

at

month 12

A

CRT at month 12

regression of polyps BCVA The proportion of patients who gained

Not available

complete regression of polyps in PDT group was significantly higher than that in IVR group at months 3, 6 and 12, but no significant difference at month 24

PT

regression of

Number of patients achieving complete

Number of Treatments

ED

complete

BCVA

CRT

Not available

achieving

Ranibizumab only

CRT

The number of patients in

Ranibizumab or bevacizumab

Number of Treatments

M

outcomes

A

anti-VEGF

N U SC R

than age-related macular degeneration.

Monotherapy was superior to combination

No significant difference between the groups in total,

Monotherapy was superior to

therapy

but BCVA had a better improvement with

combination therapy

monotherapy in SF PDT subgroup when Kaiser’s study belonged to SF PDT group.

No significant difference between the groups

No significant difference between two groups in total,

No significant difference between two

and couldn’t conclude that CRT was thinner in

but CRT was thinner with combination therapy in SF

groups.

combination therapy

PDT subgroup and CRT was thinner with monotherapy in RF PDT subgroup when Kaiser’s study belonged to RF PDT group.

Proportion of

Monotherapy was superior in the proportion of

patients

patients gaining ≥ 15 letters

Monotherapy was superior in the proportion of patients

Monotherapy was superior in the

gaining ≥ 15 letters when Kaiser’s study belongs to SF

proportion of patients gaining ≥ 3

I PDT group

letters Number of

N U SC R

gaining/losing the

Combination therapy could decrease the

No significant difference in the number of treatments

Just roughly mentioned that four studies

injections of ranibizumab

between the two groups

showed a reduction of the number of

CC E

PT

ED

M

A

injections

A

letters

ranibizumab injections in combination group without a statistical analysis.