STUDY SUBJECTS

Major inclusion criteria:

• Adults (age >18)
• Diagnosis of diabetes mellitus (either Type 1 or Type 2)
• Extensive, active, neovascular or fibrovascular proliferations- defined as 1) severe new vessels and severe fibrous proliferations, 2) severe new vessels and red vitreous hemorrhage, 3) moderate new vessels, severe fibrous proliferation and red vitreous hemorrhage
• BCVA better than 10/200

Major exclusion criteria

• Renal disease severe enough to require transplant or chronic dialysis (or other coexisting disease that would make survival or availability for at least 2 years unlikely)
• Severe NVI, NVA (involving >3 clock hours of the TM), NVG, or IOP more than 30mmHg despite treatment
• Retinal detachment involving center of the macula with visual acuity less than 10/50
• Visual acuity less than 10/200
• History of prior vitrectomy

RANDOMIZATION SCHEME AND INTERVENTIONS

Baseline visit, then subsequent randomization visit 2 weeks to 3 months later

Randomized 1:1

(a) Early vitrectomy

(b) Conventional therapy (observation with vitrectomy at specific indications during the follow-up)

RESULTS

Study population

• 381 eyes from 345 patients randomized, in 4 year analysis 181 early vitrectomy, 189 conventional management
• History of prior PRP was symmetric between groups (64.6% early vitrectomy and 66.1% conventional)

Visual acuity end-points

• Percentage of eyes with visual acuity of 10/20 or better at 48 months: 40% in early vitrectomy group versus 29% in the deferred vitrectomy group, which was significant (and was a trend that started about 6 months after enrollment)
• Stratification by severity of new vessels demonstrated that patients with more severe new vessels (advanced disease) in particular were more likely to obtain good final visual acuity (>10/20) with early vitrectomy
• Early vitrectomy was actually disadvantageous in the least severe category of PDR, compared with conventional management.

Poor Outcomes

• Percentage of eyes with NLP vision: rose rapidly in early vitrectomy group (13% at one year) compared with the deferred vitrectomy group (4%), however by 48 months there was no significant difference between these groups (21% versus 19%, respectively)

Retinal Detachment

• More frequent retinal detachment more often in the conventional management group compared with the early group by four years (37% versus 21%)

Of Note

• Patients with history of PRP prior to enrollment in the study had higher odds of finishing with better visual acuity (this was statistically significant), regardless of study arm
• 91 eyes in the conventional management group (out of 189) ultimately underwent vitrectomy during the study period, most occurred within the first 18 months of enrollment

CONCLUSIONS

• Early vitrectomy provided advantage for obtaining better visual acuities at four years, particularly in eyes with more advanced disease as graded by the amount of neovascularization and fibrovascular proliferation. Prompt vitrectomy also allowed for faster visual rehabilitation compared with conventional management, however was also associated with a higher incidence of NLP eyes (particularly early during the follow-up period).
• The authors note that they do not recommend vitrectomy as a replacement for PRP, but rather as an adjunct when there is inadequate vessel regression following PRP, or hemorrhage precludes placement of PRP