|Home | About | Journals | Submit | Contact Us | Français|
To compare the effectiveness and safety between bevacizumab and ranibizumab in the treatment of age-related macular degeneration (AMD) through a systematic review and meta-analysis.
We performed a comprehensive search of randomized controlled trials (RCTs), non-RCTs, case-control and cohort studies that compared bevacizumab and ranibizumab using PubMed and the Cochrane Library. After the related data were extracted by two investigators independently, pooled weighted mean differences (WMDs) and risk ratios (RRs) with 95% confidence intervals (CIs) were estimated using a random-effects or a fixed-effects model.
A total of four RCTs involving 1927 patients and eleven retrospective case series involving 2296 patients were included. For the primary outcomes, no significant differences were found between ranibizumab group and bevacizumab group in visual acuity (WMD: -0.04; 95%CI: -0.08 to 0.00; P=0.06), best corrected visual acuity (WMD: -0.05; 95%CI: -0.10 to 0.00; P=0.05), retina thickness (WMD: -4.69; 95%CI: -13.15 to 3.76; P=0.86) and foveal thickness (WMD: 10.91; 95%CI: -14.73 to 36.56; P=0.40). The pooled analyses in the evaluation of safety showed that compared to bevacizumab, ranibizumab was associated with decreased risks of ocular inflammation (RR: 0.45; 95% CI: 0.23 to 0.89; P=0.02) and venous thrombotic events (RR: 0.27; 95%CI: 0.08 to 0.89; P=0.03). However, there were no significant differences observed in deaths (P=0.69) and arterial thromboembolic events (P=0.71) between the two groups.
With equal clinical efficacy, ranibizumab was found to be associated with less adverse events compared to bevacizumab, indicating that ranibizumab might be a safer management.
Age-related macular degeneration (AMD) is one of the major causes of blindness in developed countries-. It is the third leading cause of blindness, coming after cataract and glaucoma, accounting for 8.7% of all legal blindness across the world. The number of individuals affected is estimated to be doubled by the year 2030 owing to the longevity of the aged population. Hence, AMD becomes a major public health problem with significant economic and social impact. Population studies indicate that neovascular AMD accounts for two thirds of late AMD cases and 90% of blindness from AMD.
Vascular endothelial growth factor (VEGF), which is regulated by hypoxia, promotes angiogenesis, and its role in the pathogenesis of neovascular AMD is well recognized,. The advent of intravitreous VEGF inhibitors has renovated the management of neovascular AMD. There are various anti-VEGF drugs commonly used nowadays, such as pegaptanib, ranibizumab and bevacizumab,. The effectiveness of pegaptanib was not as ideal as ranibizumab and bevacizumab, visual decline was still seen in the AMD patients after treatment.
Bevacizumab is a humanized anti-VEGF monoclonal IgG1 antibody,. In combination with chemotherapy, it was originally approved by the Food and Drug Administration for the treatment of various cancers, such as colorectal cancer, non-small cell lung cancer and renal cell cancer. The effectiveness of bevacizumab on wet AMD was first shown by Rosenfeld et al. Ranibizumab, a recombinant monoclonal antibody fragment that inhibits VEGF, has been approved for the treatment of all angiographic subtypes of subfoveal neovascular AMD by the Food and Drug Administration since 2006 and by the European Medicines Agency since 2007. The approval was based on two randomized clinical trials (RCTs), in which approximately 95% of the patients treated with monthly ranibizumab injections lost fewer than 15 letters in 12mo, compared to 64% of patients receiving photodynamic therapy (PDT) and 62% receiving sham treatment. The costs of ranibizumab, however, are immense. With monthly injections at a dose of 0.5 mg, the annual costs count up to more than US$23 000 per patient, about 10 times more than that of bevacizumab,. Although many studies, including large RCTs, tried to compare the efficacy of ranibizumab and bevacizumab for the treatment of AMD, the results were controversial. Therefore, a systematic review and meta-analysis of pooled data from RCTs and non-RCTs were performed in this study, aiming to evaluate the clinical effectiveness of bevacizumab and ranibizumab in the treatment of AMD.
The systematic review and meta-analysis considered RCTs and non-RCTs comparing bevacizumab versus ranibizumab for the treatment of patients with AMD. We searched PubMed (1966-October 2012) and the Cochrane Library (1988-October 2012) without language restrictions. Search terms including MeSH words and text words. The terms we used were ‘Lucentis’ or ‘ranibizumab’ or ‘Avastin’ or ‘bevacizumab’ or ‘age-related macular degeneration’ or AMD’. Furthermore, we perused the bibliographies of retrieved articles and relevant reviews. If the studies did not contain all of the necessary information, we contacted the authors directly to obtain the missing data.
For inclusion, studies had to meet the following criteria: 1) RCTs or non-RCTs studies, which compared the efficacy or safety between bevacizumab and ranibizumab in patients with AMD. Studies with full data information needed were included in the meta-analysis; 2) at least one of the primary outcomes [i.e. visual acuity (VA), best-corrected visual acuity (BCVA), foveal thickness (FT), retina thickness (RT) and central macular thickness (CMT)] or secondary outcomes (serious adverse effects, such as ocular inflammation, deaths and thromboembolic events) were evaluated; 3) enrolled a minimum of 10 eyes. If multiple papers from the same study were identified, only the one with the most detailed information and longest follow-up was selected for inclusion. Studies were excluded if they: 1) included patients with other diseases but not AMD, including choroidal neovascularization, choroid melanoma, drusen, subretinal hemorrhage and diabetic macular retinopathy; 2) evaluated bevacizumab or ranibizumab as monotherapy; 3) had no original data (reviews, comments or letters), and 4) not conducted in humans.
To avoid bias in the data extraction process, two investigators (Zhang XY and Guo XF) independently extracted and collected data following the selection criteria described above. Any discrepancy was resolved by discussion and consensus. The following information was extracted from each trial: first author's name, publication year, type of study, the number of treated patients, duration of follow-up, dosage, injections per patient and main findings. An electronic abstraction database was established in Microsoft Excel. We evaluated the quality of the studies included in this research with the Jadad score for RCTs and Newcasle-Ottawa Scale (NOS) for non-RCTs. The range of Jadad score is from 1 to 5 and the range of NOS is from 1 to 9,.
To evaluate the efficacy and safety between bevacizumab and ranibizumab for the treatment of AMD, we assessed the overall effect of bevacizumab and ranibizumab from the data of the included studies and used the weighted mean differences (WMDs) and risk ratios (RRs) with 95% confidence intervals (CIs) as the metric of choice for all the outcomes. We implemented meta-analysis of the direct evidence for each outcome, combining pairwise comparisons between bevacizumab and ranibizumab using Review Manager 5.0. Between-study heterogeneity was evaluated by Q-statistic and quantified by the I2 statistic. If statistically significant heterogeneity was considered present (P<0.1 and I2>50%), we chose a random-effects model, otherwise, a fixed-effects model was used. The value of P less than 0.05 was regarded as statistically significant for all included studies.
We identified 1545 potentially relevant studies from the initial search, and 1514 trials were excluded after a preliminary review. The remaining 31 studies were identified for detailed assessment. Finally, 4 RCTs and 10 retrospective chart series met the inclusion criteria. The selection process and reasons for exclusion are summarized in Figure 1-.
The baseline characteristics of the participants and the design of the studies are summarized in Tables 1 and and2.2. Of the 4 RCT studies, two were conducted in the United States, and two in the United Kingdom and India each. The follow-up durations in all the included studies ranged from 2 to 24mo. Of the 15 studies, with age ranging from 63 to 90y, fourteen included both genders. For the study of Subramanian, there was only male patient in group B. Two RCTs had a Jadad score of 5, and the other two had a score of 3. For the non-RCTs, one trial had a NOS score of 8, two had a score of 7, each three had a score of 5 and 6, and the remaining one trial had a score of 3. In Tables 3–6 it shows the main results from each included study for our primary and secondary outcomes.
Figure 2 shows the forest plots of 3 RCTs with 4 populations involving 1410 patients for the effect of VA. The mean difference of VA was not significant between the ranibizumab group and the bevacizumab group (WMD: -0.04; 95%CI: -0.08 to 0.00; P=0.06), with no evidence of heterogeneity (I2=0%, P=0.61). Two studies reported data for the mean BCVA. The pooled result showed that the mean BCVA was not significantly different between the two groups (WMD: -0.05; 95%CI: -0.10 to 0.00; P=0.05, data not shown).
Figure 3 presents 3 studies involving 1448 patients for the effect of RT. The overall result showed that the mean RT was not significantly thinner in the ranibizumab group than the bevacizumab group (WMD: -4.69; 95%CI: -13.15 to 3.76; P=0.86). This finding was consistent for both RCTs (WMD: -4.83; 95%CI: -13.44 to 3.78; P=0.27) and non-RCTs (WMD: -0.86; 95%CI: -45.62 to 43.90; P=0.97). The heterogeneity test was not significant (I2=0%, P=0.94).
Figure 4 shows the forest plot of 3 RCT studies and 3 non-RCT studies involving 1588 patients for the effect of FT. The overall result showed that the mean difference of FT was not significant between ranibizumab group and bevacizumab group (WMD: 10.91; 95%CI: -14.73 to 36.56; P=0.40), with a significant heterogeneity (I2=84%, P<0.0001). Subgroup analyses showed that the result was consistent in both RCTs and non-RCTs.
Figure 5 shows the forest plot comparing the safety between ranibizumab and bevacizumab. In the pooled result of 3 RCTs and 1 non-RCT, more patients died in bevacizumab group compared to ranibizumab group. However, this difference was not statistically significant (RR: 0.92; 95%CI: 0.62 to 1.38; P=0.69; Figure 5A), with significant heterogeneity (I2=0%, P=0.88). The overall result from 3 RCTs and 3 non-RCTs showed that ranibizumab was not associated with a reduction in the risk of arterial thromboembolic events (RR: 0.75; 95%CI: 0.16 to 3.42; P=0.71; Figure 5B), with consistent result in both RCTs and non-RCTs. The risk of ocular inflammation was reported in 2 RCTs and 7 non-RCTs. The overall result showed that ranibizumab was associated with a decreased risk of ocular inflammation compared to bevacizumab (RR: 0.45; 95%CI: 0.23 to 0.89; P=0.02; Figure 5C), without heterogeneity (I2=45%, P=0.11). However, this finding was only significant in non-RCTs (RR: 0.40; 95%CI: 0.18 to 0.91; P=0.03). Figure 5D shows the forest plot of venous thrombotic events from 2 RCTs involving 1795 patients. The risk of venous thrombotic events was significantly less in the ranibizumab group than the bevacizumab group (RR: 0.27; 95%CI: 0.08 to 0.89; P=0.03). The heterogeneity test was not significant (I2=0%, P=0.79). Five studies investigated the serious ocular adverse, with four of them having no events in both groups and one RCT [CATT 2012] indicating that the risk was lower in the ranibizumab group (RR: 0.79; 95%CI: 0.68 to 0.93; P=0.03, data not shown).
The studies included in this system review indicate robust efficacy and safety from ranibizumab and bevacizumab treatment based on RCTs and non-RCTs. The results of our meta-analysis suggest that ranibizumab and bevacizumab have equal clinical efficacy. However, the pooled analyses in the evaluation of safety showed that compared to bevacizumab, ranibizumab was associated with decreased risks of ocular inflammation and venous thrombotic events.
Although some systematic reviews investigated the efficacy and safety of ranibizumab and bevacizumab in AMD, the outcomes were assessed separately rather than a direct comparison and the conclusions were based on descriptive analysis. In the present study, we included studies that compared the two drugs directly and found that the VA, RT and CFT of ranibizumab in the treatment of AMD were, at least, equivalent to those of bevacizumab.
The epitopes of ranibizumab and bevacizumab locate in the receptor-binding region of VEGF, and both antibodies target VEGF in a similar way. However, bevacizumab (149 kDa) and ranibizumab (48.39 kDa) have different molecular weights, mainly because ranibizumab does not contain an Fc part. Moreover, bevacizumab is produced in a eukaryotic cell line and is N-glycosylated in its Fc region, but ranibizumab is expressed in prokaryotic E. coli without any glycosylation sites. Therefore, the various molecular mechanisms of the drugs might result in different efficacy. Debates remained in the past years on whether ranibizumab or bevacizumab is superior in treating AMD. Chang et al argued that being a smaller molecule, it is easier for ranibizumab to permeate the retina and inhibit abnormal blood vessel growth, thus leading to a better short-term efficacy of ranibizumab compared to bevacizumab. On the contrary, bevacizumab was found to be superior in long-term effects because of its decreased clearance from eye due to the larger size, and the consequent high accumulation in retinal pigment epithelial (RPE) cells. In our study, no difference was observed between ranibizumab and bevacizumab in terms of efficacy, likely that many mechanisms interplay in the clinical practice and the management is perhaps more complicated than we assumed. More standard clinical trials are needed to be done to conclude superiority.
Recently, intravitreal anti-VEGF drug injection has been reported with complications and adverse events, mainly including subconjunctival hemorrhage, cornea tear, ocular inflammation, uveitis, retinal detachment and cataract,. Some studies compare PDT with either intravenous ranibizumab or bevacizumab-. Intravitreal injection of ranibizumab was reported to be associated with endophthalmitis (≤2.1%), uveitis (≤1.3%), retinal detachment (≤1.5%), retinal tear (≤1.9%) and vitreous hemorrhage (≤8.0%)-. Compared to PDT group, an increase rate of pigment epithelial tears (5.5% vs 0.0%), posterior vitreous detachment (14.6% vs 0.0%) or cataract progression (7.3% vs 0.0%) was found in bevacizumab group in one RCT. Although many studies assessed the safety of ranibizumab or bevacizumab comparing to control group, the comparison was not direct and likely led to an inconclusive result. In a previous meta-analysis, Schmucker et al found that the difference of arterial thromboembolic events, serious nonocular hemorrhage and death were not statistically significant between the two drugs. But a pooled analysis of serious ocular adverse events indicated a significantly increased RR for bevacizumab when compared to ranibizumab. In combination of 2-year follow-up result of CATT study and the new RCT IVAN trial, we found a higher risk of bevacizumab in ocular inflammation and venous thrombotic events, indicating a better safety profile of ranibizumab in AMD patients,. There were no substantial imbalances in demographic or ocular characteristics at baseline, indicating that the increased incidence of venous thrombosis is the result of truly higher risk. Regarding the safety profile of the two drugs, a previous meta-analysis including 11 studies, found that an increased risk of ocular and multiple systemic ocular adverse effects with bevacizumab, strengthening the better safety profile of R. With equal efficacy and better safety profile compared to bevacizumab, ranibizumab seems to be the prior choice of AMD. However, the issue of expensiveness remains with ranibizumab.
Additionally, in some studies the effect of ranibizumab and bevacizumab on retinal conditions was compared. Singer et al concluded that in patients with retinal vein occlusions, ranibizumab appeared to have a greater short-term effect in decreasing macular edema on OCT when compared to bevacizumab. In another study by Niederhauser et al, the effect of bevacizumab or ranibizumab on visual acuity and central foveal thickness was evaluated in macular edema also resulted from retinal vein occlusion. However no significant differences between bevacizumab and ranibizumab were found in the study.
The present study had several limitations. First, the publication bias cannot be fully ruled out. The number of studies included is insufficient to carry out a further statistical analysis to detect publication bias through asymmetry plot. Second, the studies included were heterogeneous in terms of study location, population and basal condition. We were not able to use individual-level data to improve the quality of adjustment and the precision of estimates. Finally, the delay between literature search and publication was inevitable.
Conflicts of Interest: Zhang XY, None; Guo XF, None; Zhang SD, None; He JN, None; Sun CY, None; Zou Y, None; Bi HS, None; Qu Y, None, Wang HL, None; Li RX, None.