Protocol S, Five-Year Results
Five-Year Outcomes of Panretinal Photocoagulation vs. Intravitreous Ranibizumab for Proliferative Diabetic Retinopathy
Citation: Gross JG, Glassman AR, Liu D, Sun JK, Antoszyk AN, Baker CW, Bressler NM, Elman MJ, Ferris FL 3rd, Gardner TW, Jampol LM, Martin DF, Melia M, Stockdale CR, Beck RW; Diabetic Retinopathy Clinical Research Network. Five-Year Outcomes of Panretinal Photocoagulation vs Intravitreous Ranibizumab for Proliferative Diabetic Retinopathy: A Randomized Clinical Trial. JAMA Ophthalmol. 2018 Oct 1;136(10):1138-1148. doi: 10.1001/jamaophthalmol.2018.3255. Erratum in: JAMA Ophthalmol. 2019 Apr 1;137(4):467. PMID: 30043039; PMCID: PMC6233839.
- Protocol S was a multicenter, randomized clinical trial of patients with proliferative diabetic retinopathy (PDR) that compared treatment with intravitreal ranibizumab to pan-retinal photocoagulation (PRP).
- At 5 years, severe vision loss and serious PDR complications were uncommon with both ranibizumab and PRP, but the ranibizumab group had lower rates of vision-impairing diabetic macular edema and less visual field loss.
- These findings support either anti-vascular endothelial growth factor therapy or PRP treatment for patients with PDR.
To evaluate efficacy and safety of 0.5-mg intravitreous ranibizumab versus PRP over 5 years for PDR.
Multicenter, randomized, clinical trial
- Patients with PDR and best-corrected visual acuity (BCVA) letter score of 24 or better.
Major inclusion criteria:
- Patients who previously underwent PRP
Major exclusion criteria:
RANDOMIZATION SCHEME AND INTERVENTIONS
Randomized 1:1 to either 0.5-mg intravitreous ranibizumab or PRP. If participants had 2 study eyes, they randomly received ranibuzumab in one eye and PRP in the other.
Eyes in the ranibizumab group received a baseline injection and monthly injections through week 24. At weeks 16 and 20, deferral of injections was allowed if neovascularization (NV) was resolved. Starting at week 24, injections could be deferred if all NV was resolved or if NV became and remained stable after 2 consecutive injections. If NV worsened, injections were resumed. If an eye met failure or futility criteria, PRP was permitted.
Eyes in the PRP group received PRP—either full or completed over several sittings—at baseline. Additional PRP was performed if the amount of NV increased.
Both groups required ranibizumab injections at baseline for diabetic macular edema (DME). Vitrectomy for vitreous hemorrhage was not allowed for the first 8 weeks; vitrectomy for retinal detachment was allowed at investigator discretion.
- The study initially enrolled 394 study eyes (n=191 for ranibizumab, n=203 for PRP) from 305 patients overall
- The 5-year visit was completed for 240 study eyes (61%) in the ranibizumab (117 [69%]) and PRP (123 [65%]) groups.
- Mean (SD) age was 52 (12) years, 135 (44%) were women, and 160 (52%) were white.
- Ranibizumab Group: over the 5 years of study, the mean (SD) number of injections were 7.1 (2.2), 3.3 (2.9), 3.0 (2.8), 2.9 (2.7), and 2.9 (2.8), respectively. The number of eyes not receiving an injection during years 2 through 5 were 29 (25%), 32 (27%), 33 (28%), and 43 (37%), respectively. Focal/grid laser treatment was required for 16 (8%) eyes, and PRP was required for 26 (14%) eyes (with 18 of these being endolaser during vitrectomy).
- PRP Group: after the initial baseline PRP treatment, at least 1 additional PRP session was administered in 103 (51%) eyes, with most occurring within the first 6 months (n=39; 38%) or within 6-12 months (n=28; 27%). During the 5 years, 71 (58%) eyes received at least 1 ranibizumab injection for DME. Focal/grid laser treatment was performed in 26 (13%) eyes.
- Change in Baseline Visual Acuity: At the 5-year mark, visual acuity improved a mean (SD) of 3.1 (14.3) letters in the ranibizumab group and 3.0 (10.5) letters in the PRP group. This translates to a mean visual acuity (approximate Snellen equivalent) of 20/25 in each group, compared to 20/32 at baseline. These results did not differ significantly between the two groups (P=.68)
- At the 5-year visit, among eyes with baseline DME (n=40), the mean (SD) central subfield thickness (CST) decreased from baseline by 139 (157) μm in the ranibizumab group and 59 (102) μm in the PRP group. This difference was not statistically significant (P=.16). At 5 years, CST increased from baseline by at least 10% in 5 (5%) of ranibizumab eyes and 14 (14%) of PRP eyes.
- The mean (SD) change in cumulative visual field total point score from baseline, combining Humphrey visual field 30-2 and 60-4 test patterns, was -330 (645) dB in the ranibizumab group and -527 (635) dB in the PRP group; the PRP group had significantly lower cumulative visual field total point scores (P=.04).
- Retinal detachment was identified in 12 ranibizumab eyes and in 30 PRP eyes (P=.004), with most (10 and 24, respectively) being tractional retinal detachments. Vitrectomy was performed for retinal detachment in 4 ranibizumab eyes and 11 PRP eyes.
- Vitreous hemorrhage developed in 91 ranibizumab eyes and 93 PRP eyes (P=.47).
- No differences were identified with a P value less than .01 among these 3 groups for percentage of participants with serious adverse events, hospitalizations, deaths, or events in each MedDRA system organ class (except for Infections and Infestation, 39 [44%] in the bilateral group, 39 [38%] in the ranibizumab group, and 24 [21%] in the PRP group [P=.001]).
- Arteriothrombroembolic events were reported in 12 of 89 participants (13%) with 2 study eyes. For those with 1 study eye, events were reported in 18 of 102 (18%) ranibizumab participants and 12 of 114 (11%) PRP participants. These were not statistically significant differences (P=.31).
- Injection-related endophthalmitis occurred in 1 of 4113 injections.
- These findings support either ranibizumab or PRP as viable treatments for PDR.
- For those that completed follow-up, visual acuity was very good over 5 years and occurrence of severe vision loss or serious complications of PDR were uncommon with either PRP or ranibizumab.
- Ranibizumab resulted in lower rates of development of vision-impairing DME and less visual field loss.
- There were no meaningful differences in patient-centered outcomes, and so individual patient factors should be considered when choosing a treatment for PDR.