Factors affecting anatomical and visual outcomes in Terson syndrome managed by surgery or observation

Objectives: To investigate visual outcomes and prognostic factors of patients with Terson syndrome (TS) managed with observation or pars plana vitrectomy (PPV).

Methods: retrospective review of medical records of 117 eyes from 81 patients (43 females) with TS. Main outcome measures were final best corrected visual acuity (BCVA), intraoperative findings and long-term sequelae.

Results: 46 (39.3%) eyes were managed conservatively and 71 (60.7%) eyes underwent PPV. Median follow-up was 8.4 months. The PPV group had significantly worse (p < 0.001) baseline BCVA (median 2.3 versus 0.7 logMAR, Snellen equivalent 20/4000 versus 20/100). Final BCVA did not differ between the two groups (p = 0.38). Final BCVA ≥ 0.3 logMAR (20/40) in the surgery group was associated with post-operative retinal detachment (p = 0.013) and macular abnormalities (p = 0.014), and in the observation group with ocular comorbidity (p = 0.008). Retinal breaks were detected intraoperatively in 25 (35.2%) eyes and were associated with an interval longer than 3 months between ocular diagnosis and surgery (p = 0.04), but not with larger gauge instrumentation and posterior vitreous detachment. Incidence of ERM did not differ among patients managed conservatively and after PPV (p = 0.9) and between eyes undergoing early or delayed surgery (p = 0.09). The most common post-operative complications were cataract in 16 (22.5%) eyes and ERM in 8 (11.3%) eyes.

Conclusions: visual outcomes in TS are similar with both management strategies. Surgery allows faster and greater visual recovery but carries high risk of intraoperative retinal tears if delayed for longer than 3 months from initial presentation. ERM and retinal detachment are not correlated with timing of surgery or management strategy.