Ranibizumab entered clinical trials for neovascular age-related macular degeneration (nAMD) in 2000, an investigative sequence that culminated with the publication of the ANCHOR  and MARINA  trials in 2006 and approval by the US Food and Drug Administration (FDA) in the same year. At approximately the same time, two small pilot studies explored the safety and efficacy of ranibizumab in patients with DME. In the READ-1 study , ten patients with chronic DME received intravitreal injections of 0.5 mg ranibizumab at baseline and 1, 2, 4, and 6 months. By the 7-month endpoint, mean best-corrected visual acuity (BCVA) had improved by +12.3 letters. Mean macular volume improved from 9.22 to 7.47 mm2 (77% reduction in excess volume) and mean foveal thickness improved from 503 µm to 257 µm (85% reduction in excess thickness). The other study enrolled ten patients in an open-label, dose-escalation protocol . Patients received low-dose (0.3 mg) or high-dose (0.5 mg) ranibizumab at baseline, and 1 and 2 months. The mean improvements in BCVA and macular thickness at 3 months for the 0.3-mg and 0.5-mg arms were +12 and +7.8 letters and −45.3 and −197.8 µm, respectively. Improvements in BCVA decreased slightly between 3 and 6 months.
These studies demonstrated that DME responds well to ranibizumab, but they failed to demonstrate superiority of either the 0.3-mg or 0.5-mg dose and suggested that 2-month injection intervals may be sufficient. Subsequent development of ranibizumab for DME proceeded along three research lines: READ-2/RISE and RIDE (Genentech-sponsored in the USA and South America); RESOLVE/RESTORE (Novartis-sponsored ex-US); and Diabetic Retinopathy Clinical Research Network (DRCR.net) Protocol I (National Eye Institute-sponsored). Table 2 lists the important DR clinical trials with ranibizumab and details their key findings.
READ/RISE and RIDE
Following successful completion of the READ-1 trial, the phase II READ-2 trial was performed at 14 sites . This prospective controlled trial randomized 126 patients to receive 0.5 mg ranibizumab at baseline and 1, 3, and 5 months (group 1), laser at baseline and at month 3 if needed (group 2), and ranibizumab combined with laser at baseline and month 3 if needed (group 3). At the 6-month primary endpoint, the improvement in BCVA for group 1 was significantly greater than for group 2 (+7.24 vs. −0.43 letters; P = 0.0001) but not significantly different from that of group 3 (+3.8 letters). Fifteen-letter improvement in BCVA was achieved by 22%, 0%, and 8%, and excess macular thickness was reduced in 50%, 33%, and 45%, respectively.
Between months 6 and 24, all patients became eligible to receive ranibizumab injections every 2 months. At 24 months, mean improvements in BCVA were +7.7, +5.1, and +6.8 letters, and the proportions of patients improving by 15 letters were 24%, 18%, and 26%. Mean central subfield thickness (CST) measurements were 340, 286, and 258 µm. By month 24, mean visual acuity improvement experienced by the cohort initially randomized to laser approached that of the ranibizumab group, but the excess macula thickness in the ranibizumab monotherapy group indicated that these patients were probably under-treated .
At 24 months, the protocol was amended a second time to allow all patients to receive monthly pro re nata (PRN) ranibizumab in year 3. From the 24-month to the 36-month visits, patients had mean BCVA gains of +3.1, −1.6, and +2.0 letters. An average of 5.4, 2.3, and 3.2 ranibizumab injections were administered during the third year .
A higher dose of ranibizumab (2.0 mg) had been tested during the exploratory dosing nAMD trials, but it was quickly abandoned because of its high viscosity and tendency to cause inflammation. Several years later, however, the 2.0-mg dose was re-evaluated in the HARBOR trial  for nAMD and the READ-3 trial  for DME. Patients with DME received either the 2.0-mg or the 0.5-mg ranibizumab dose monthly for 6 months, followed by monthly PRN injections through month 12. Patients receiving 0.5 mg experienced a non-statistically significant greater improvement in BCVA (+9.34 letters vs. +7.04 letters), and more patients receiving the 2.0-mg dose died from myocardial infarction (4% vs. 1%). Because neither the READ-3 nor HARBOR trial showed that high-dose ranibizumab offered a clinical advantage, further development of the 2.0-mg dose was halted. The READ trials suggested that monthly ranibizumab was needed to produce optimal visual acuity results, but that higher doses provided no added benefit and might actually increase the risk of serious systemic adverse events.
As a result, the multi-center, double-masked parallel 3-year phase III RISE and RIDE registration trials compared the efficacy of lower-dose (0.3 mg and 0.5 mg) monthly ranibizumab against sham for patients with center-involving DME . Seven hundred fifty-nine patients were randomized to one of three treatment arms: monthly 0.3 mg ranibizumab, monthly 0.5 mg ranibizumab, or sham. Patients were eligible for rescue laser photocoagulation at 3 months if the central retinal thickness (CRT) was >250 µm and if the change in CRT following the previous injection was <50 µm. The primary endpoint was the proportion of subjects improving by at least +15 letters at 24 months, and secondary endpoints included improvement in BCVA, improvement in macular thickness, and safety measures.
At the 24-month primary endpoint, the proportion of patients who improved by at least +15 letters was significantly greater in patients receiving 0.3 mg and 0.5 mg ranibizumab than in patients who received sham injections in both RISE (44.8%, 39.2%, and 18.1%) and RIDE (33.6%, 45.7%, and 12.3%). Mean improvement in BCVA was also greater in the groups receiving ranibizumab in both RISE (+12.5, +11.9, and +2.6 letters) and RIDE (+12.0, +10.9, and +2.3 letters), as were corresponding mean improvements in central foveal thickness (CFT) in RISE (−250.6 µm, −253.1 µm, and −133.4 µm) and RIDE (−259.8 µm, −270.7 µm, and −125.8 µm). Fewer patients receiving ranibizumab required rescue laser (means of 0.3 to 0.8) compared to those randomized to sham injections (means of 1.8 and 1.6). In the groups treated with 0.3 mg ranibizumab, 37.2% experienced a two-step or greater improvement in diabetic retinopathy severity score (DRSS), and 13.2% experienced a three-step or greater improvement.
The median DRSS remained stable at moderately severe NPDR in the sham groups, whereas scores improved from moderately severe to mild NPDR in the ranibizumab groups. Fewer patients receiving ranibizumab experienced a two-step worsening in the DRSS (1.7% to 2.1% vs. 9.6% in sham), and fewer ranibizumab patients developed vitreous hemorrhage. Since changes in BCVA were the same for patients receiving the 0.3-mg and 0.5-mg doses of ranibizumab, and the incidence of several categories of adverse events including central nervous system and cerebrovascular hemorrhage appeared to increase in a dose-dependent fashion (although the numbers were low), Genentech recommended the 0.3-mg dose for approval. The US FDA ultimately approved the use of 0.3 mg ranibizumab for DME on August 10, 2012. The label was subsequently expanded (2015) to include the treatment of diabetic retinopathy (NPDR and PDR) in patients with DME.
During year 3, patients continued receiving monthly ranibizumab injections, and those randomized to sham were eligible to receive monthly 0.5 mg ranibizumab . From the 24-month to the 36-month visits, patients in the ranibizumab arms had stable BCVA, whereas those in the sham arms improved to +4.3 (RISE) and +4.7 (RIDE) letters above baseline.
After the 36-month visit, 500 patients were followed in the extension study for a mean of 14.1 months . All patients were eligible to receive monthly PRN 0.5 mg ranibizumab if DME was identified by the investigator or BCVA worsened by at least five letters compared to month 36. Patients received a mean of 4.5 injections (annualized: 3.8), but 24.2% of patients did not require any injections. BCVA remained stable in all groups and mean CFT increased slightly. Few patients developed PDR, and those originally randomized to ranibizumab had a lower overall rate of progression to PDR than those originally randomized to sham.
The phase II RESOLVE and phase III RESTORE trials that were performed in Europe, Asia, and Australia served as the basis for regulatory approval of ranibizumab in these regions for treatment of DME. In the 12-month RESOLVE trial, 151 patients were randomized to receive monthly 0.3 mg ranibizumab, 0.5 mg ranibizumab, or sham injections . After 1 month, the drug dose could be doubled if the CRT was greater than 300 µm, or greater than 225 µm if the CRT decreased by <50 µm following the previous injection. At 3 months, patients received ranibizumab injections or sham injections on a monthly PRN basis, and all were eligible for rescue laser photocoagulation. At the 12-month primary temporal endpoint, mean BCVA improved by +10.3 letters in the pooled ranibizumab groups but declined by −1.4 letters in the sham group. BCVA gains of +2 lines and +3 lines were achieved by 60.8% and 32.4% of ranibizumab-treated eyes, but only 18.4% and 10.2% of sham-treated eyes, respectively. Mean improvements in CRT were −194.2 µm for the pooled ranibizumab groups and −48.4 µm for the sham group. Eighty-six percent of all eyes received a ranibizumab dose of 0.5 to 1.0 mg at some point during the trial. The mean number of administered injections was 10.2, and only 4.9% of ranibizumab treated-eyes (compared to 34.7% of sham eyes) required rescue laser.
Since nearly all patients in the RESOLVE trial eventually received a ranibizumab dose of at least 0.5 mg, only the larger (0.5 mg) dose was included in the design of the RESTORE trial. RESTORE was the multi-center (73 sites) phase III ranibizumab registration trial for the Eastern Hemisphere . Three hundred forty-five patients were randomized to receive ranibizumab + sham laser, ranibizumab + laser, or sham injections + laser. Ranibizumab injections were administered monthly for three months, then monthly PRN; laser was performed at baseline and then every 3 months PRN. The trial’s primary objective was to demonstrate superior mean BCVA improvement at 12 months in patients treated with ranibizumab monotherapy or ranibizumab combined with laser compared to laser monotherapy. Secondary objectives included the proportions of patients achieving BCVA of at least 73 letters (20/40), the time course of the change in mean BCVA and CRT, patient-reported outcomes relative to laser photocoagulation, and safety measures.
At the 12-month primary endpoint, patients in the ranibizumab monotherapy, ranibizumab + laser, and sham/laser groups demonstrated improvements in mean BCVA (+6.1, +5.9, and +0.8 letters), BCVA score >73 letters (53.0%, 44.9%, and 23.6%) and CRT (−118.7 µm, −128.3 µm, and −61.3 µm). Health-related quality-of-life scores (measured by the 25-item National Eye Institute Visual Function Questionnaire [NEI VFQ-25]) improved more in the ranibizumab monotherapy and ranibizumab + laser groups than in the sham/laser group (P < 0.05 for each). Subgroup analyses showed that patients with baseline BCVA of ≥73 Early Treatment Diabetic Retinopathy Study (ETDRS) letters or CRT <400 µm achieved final BCVA that was as good with laser as with ranibizumab. Patients received a mean of seven ranibizumab/sham injections.
After the 12-month primary endpoint, 240 patients were enrolled in the 24-month extension trial. All patients were eligible to receive 0.5 mg ranibizumab according to BCVA and disease progression criteria, and at the investigators’ discretion. Additional laser photocoagulation was allowed according to ETDRS guidelines. At the pre-planned 24-month interim analysis, patients who originally received ranibizumab monotherapy and ranibizumab + laser maintained improvements in mean BCVA (+7.9 letters, +6.7 letters), CRT (−140.6 µm, −133.0 µm), and NEI VFQ-25 composite scores (5.6, 5.8) . Between the 12- and 24-month examinations, patients originally treated with sham/laser experienced significant improvement in these measures (+5.4 letters, −126.6 µm, 4.3) after receiving ranibizumab. Similar numbers of injections were performed in each treatment arm (3.9, 3.5, and 4.1).
Most of the patients (208; 86.7%) were rolled into and completed the 24-month extension study. While patients received as-needed ranibizumab injections during the period between the 12- and 36-month examinations, BCVA and CRT were maintained . At 36 months, improvements in BCVA were +8.0 letters (ranibizumab monotherapy), +6.7 letters (ranibizumab + previous laser), and +6.0 letters (original sham/laser arm). The mean numbers of injections over the final 2 years varied from 6.0 to 6.8 for each of the three treatment arms.
DRCR.net Protocol I
The DRCR.net Protocol I trial provided the first level I evidence that supported the use of ranibizumab as primary treatment for DME. Protocol I was a 5-year multi-center study that randomized 854 eyes with center-involving DME to receive 0.5 mg ranibizumab with prompt macular laser photocoagulation, 0.5 mg ranibizumab with deferred laser (for at least 6 months), intravitreal triamcinolone with prompt laser, or sham injections with prompt laser . Patients received ranibizumab injections during the first year according to the 4:2:7 rule—four monthly injections, followed by two additional injections if fluid persisted, followed by seven visits during which the drug could be administered at the investigator’s discretion if insufficient improvement was noted. Laser photocoagulation of the macula and intravitreal triamcinolone (4 mg) injections could be repeated every 16 weeks and 13 weeks, respectively, as needed. Patients randomized to the ranibizumab + deferred laser group were not required to receive laser treatment at 6 months and beyond if the macula was dry.
At the 12-month examination, mean improvements in BCVA in the ranibizumab + prompt laser, ranibizumab + deferred laser, triamcinolone + laser, and sham + laser groups were +9, +9, +4, and +3 letters, respectively. Most of the gains in BCVA occurred by the 8-week visit. During the first 3 months of the trial, patients who were treated with triamcinolone + laser experienced improvement in BCVA that were similar to those receiving ranibizumab. However, BCVA in the triamcinolone group worsened over the subsequent 9 months because of the development of corticosteroid-induced cataracts. On the other hand, patients in the triamcinolone/laser group who were pseudophakic at baseline experienced 1-year improvement in BCVA comparable to that in the ranibizumab arms. The following subgroup analyses found no significant differences in 1-year visual acuity outcomes: prior treatment for DME, baseline visual acuity, baseline CST, and baseline severity of DR. Improvements in CST were comparable in the groups receiving ranibizumab and triamcinolone, all of which exceeded those in the group receiving sham/laser. Eyes treated with ranibizumab were less likely to experience increases in DRSS.
During the second year of the trial, the interval between examinations could be extended to 8 weeks if treatment had been deferred for two consecutive visits, and to 16 weeks if treatment was not performed at the 8-week visit. Patients in the triamcinolone/laser and laser/sham groups were eligible to receive ranibizumab as early as week 74 for persistent edema without improved BCVA. The 2-year BCVA outcomes were similar to those at 1 year, in that 44–49% of ranibizumab-treated eyes improved by at least +10 letters and 28–29% improved by at least +15 letters . Compared to the sham/laser group, the mean changes in BCVA in patients receiving ranibizumab + prompt laser, ranibizumab + deferred laser, and triamcinolone + prompt laser were +3.7, +5.8, and −1.5 letters. Within the first 2 years of the trial, 43 eyes in the sham/laser group were switched to ranibizumab because of treatment “failure,” whereas only one of the patients randomized to ranibizumab required switching.
At the 3-year visit, the median numbers of injections given to patients in the ranibizumab + prompt laser and ranibizumab + deferred laser groups were 12 and 15, respectively , and the median numbers of laser treatments were 3 and 0, respectively. Only 46% of patients in the ranibizumab + deferred laser group had been treated with laser. Patients randomized to ranibizumab + deferred laser had BCVA improvements that were +2.9 letters greater than those randomized to ranibizumab + prompt laser (P = 0.02). The proportions of eyes with CST ≥250 µm were 36% in each of the ranibizumab arms.
At the 5-year visit, the mean BCVA improvements from baseline were +7.2 letters in the ranibizumab + prompt laser group and +9.8 letters in the ranibizumab + deferred laser group (P = 0.09) . Visual acuity loss of ≥10 letters was experienced by 9% and 8% of patients, improvement of ≥10 letters by 46% and 58%, and improvement of ≥15 letters by 27% and 38%, respectively. Only 44% of patients who were randomized to ranibizumab + deferred laser received laser photocoagulation through 5 years. The median numbers of ranibizumab injections administered during the trial were 13 and 17, 54% and 45% of patients did not receive ranibizumab during year 4, and 62% and 52% did not receive ranibizumab during year 5.
The previously discussed trials demonstrated the efficacy of both monthly and PRN therapy for DME, but most physicians treat chorioretinal vascular conditions according to a treat-and-extend (T&E) regimen. The RETAIN trial randomized 372 patients to receive T&E 0.5 mg ranibizumab + laser (G1), T&E 0.5 mg ranibizumab (G2), or monthly PRN 0.5 mg ranibizumab (G3) . Patients received monthly injections until BCVA stabilized, after which patients in G1 and G2 could be extended at 1-month intervals up to a maximum of 3 months. At the 24-month endpoint, median BCVA changes in the G1, G2, and G3 groups were +8.3, +6.5, and +8.1 letters, respectively. Seventy percent of patients receiving T&E were extended to a treatment interval of at least 2 months, and they required 40% fewer clinic visits than those receiving PRN injections.