Late-onset Ocular Hypertension after Pars Plana Vitrectomy: Five-Year Results of a Multicenter Chart-Review Study

Purpose : Purpose: To determine the incidence and risk factors of late-onset ocular hypertension after vitrectomy. Methods : Methods: In this retrospective multicenter cohort study all consecutive patients who underwent primary vitrectomy with at least 6 months, from January 2010 to December 2015, at 5 tertiary vitreoretinal centers in Italy were enrolled. Enrolled eyes were divided into 2 groups on the basis of presence of late-onset ocular hypertension after vitrectomy, determined by all of the following criteria: (1) IOP > 21 mmHg detected more than 2 months postoperatively on at least 2 separate visits; (2) an increase in IOP > 4 mmHg above the fellow eye; all eyes that developed late-onset ocular hypertension were identified as the late-onset ocular hypertension group. All other eyes that underwent vitrectomy in the same period, without late-onset ocular hypertension, were considered the control group. Patient demographics, systemic, ophthalmic, operative and postoperative data were drawn from the electronic medical records. Results : Results: From a total of 6048 vitrectomy procedures, 294 cases of late-onset ocular hypertension (4.9%) were identified in the eyes that underwent vitrectomy, and 116 eyes in the fellow eye (1.9%). Multivariable logistic regression showed that significant risk factors for developing late-onset ocular hypertension included intraoperative triamcinolone use (odds ratio [OR], 7.78; P < 0.001), longer axial length (OR, 1.55; P = 0.023), preoperative higher IOP (OR, 1.81; P = 0.003), and pseudophakic/aphakic status (OR, 2.04; P < 0.001). Decision-tree analysis showed that the stronger predictor of late-onset ocular hypertension was intraoperative triamcinolone use (P < 0.001). Secondary predictors were a preoperative IOP more than 15 mmHg (P < 0.001) in eyes that use triamcinolone, and pseudophakic/aphakic status (P = 0.007) in eyes that not use triamcinolone. Finally, the third related classification variable was an axial length >24.9 mm in eyes with pseudophakic/aphakic status (P < 0.001) that not use triamcinolone. Conclusions : Conclusions: Late-onset ocular hypertension occurs in 4.9% of vitrectomized eyes. The main risk factors are intraoperative use of triamcinolone and pseudophakic/aphakic status.