Visual Outcomes after Vitrectomy for Terson Syndrome Secondary to Traumatic Brain Injury.

PURPOSE: To evaluate visual outcomes after vitrectomy for intraocular hemorrhages secondary to traumatic brain injury. DESIGN: Retrospective, observational case series. PARTICIPANTS: A total of 28 eyes in 20 patients undergoing vitrectomy for Terson syndrome secondary to traumatic brain injury between 1997 and 2015. METHODS: We reviewed the records of patients undergoing a standard 20-gauge or 23-gauge pars plana vitrectomy for intraocular hemorrhages secondary to traumatic brain injury, and the timing of vitrectomy in relation to the inciting intracranial event was recorded. MAIN OUTCOME MEASURES: The primary outcome measure was the change in the preoperative visual acuity score at postoperative month 1 and at the last noted clinic appointment. RESULTS: A total of 28 eyes in 20 patients (all male) underwent pars plana vitrectomy for intraocular hemorrhages secondary to traumatic brain injury. The mean preoperative baseline logarithm of the minimum angle of resolution (logMAR) (Snellen) best-corrected visual acuity (BCVA) was 1.81±0.56 (20/1290). At 1-month postoperative follow-up, the mean BCVA was 0.30±0.33 (20/40). At the date of the last follow-up, the mean BCVA was 0.15±0.24 (20/30) and the median BCVA was 0.00 (20/20). Although the difference between preoperative and postoperative BVCA was significantly different at 1 month and the final postoperative clinic visits (P < 0.001), there was not a correlation between preoperative visual acuity as a predictor of final postoperative visual acuity outcome (r=-0.32; P = 0.09; 95% confidence interval [CI] -0.62 - 0.06). At the date of the last follow-up, the differences in visual outcomes between the individuals undergoing vitrectomy within 3 months of the inciting event, 0.08±0.15 (20/25), were not significantly different than those undergoing surgical intervention after 3 months, 0.18±0.27 (20/30) (P = 0.28). Three cases among those undergoing vitrectomy after 3 months were complicated by retinal detachment, none of which resulted in a BCVA worse than when the patient originally presented preoperatively. CONCLUSIONS: In this retrospective series of patients without other ocular pathology, surgical intervention effectively provided rapid visual recovery in the majority of individuals with intraocular hemorrhages secondary to traumatic brain injury, irrespective of the timing of vitrectomy or of preoperative visual acuity.