Scleral Collapse Following Phacoemulsification in Vitrectomized Eyes

Editor, Visually significant cataract develops in the majority of phakic eyes within 2 years of pars plana vitrectomy (PPV) (Braunstein & Airiani 2003). Altered fluid dynamics during phacoemulsification in these eyes can lead to excessive deepening of the anterior chamber with wide mydiasis and reverse pupil block (Cionni et al. 2004), whereas misdirection of irrigation fluid behind the iris can lead to paradoxical miosis (Ahfat et al. 2003; Braunstein & Airiani 2003). We report a novel phenomenon of posterior segment collapse noted after apparently uncomplicated phacoemulsification and discuss a possible mechanism. Three eyes were identified from a consecutive case series of 313 eyes that underwent clear corneal phacoemulsification following PPV. The surgical technique was similar throughout, with either sub-Tenon or topical anaesthesia. A deep anterior chamber was typically managed by angulation of the phacoemulsification probe and reduction of the infusion pressure. Paradoxical miosis was managed by retracting the tip of the infusion probe into the anterior chamber, or by filling the anterior chamber with viscoelastic. High infusion pressure was maintained during cortical aspiration to stabilize the capsule. After lens insertion, the anterior chamber was deepened with the infusion, the ocular pressure checked by palpation, the wound hydrated and then sutured only if there was a leak. All three cases had axial myopia (axial lengths 26.8 mm, 27.9 mm and 29.9 mm) compared with the unaffected eyes (n = 310; mean axial length 25.1 mm, median axial length 24.7 mm). All had posterior displacement of the lens/iris diaphragm during surgery but no abnormality of the red reflex was noted. Postoperative review was performed within 1 hour of surgery. In each case the anterior chamber was deep without wound leak, with intraocular pressure measurements of 7 mmHg, 14 mmHg and 4 mmHg, respectively. Unexpectedly poor visual acuity (VA) (< 6/60) prompted fundus examination that showed posterior segment collapse (Fig. 1). Ultrasonography excluded suprachoroidal haemorrhage (Fig. 2). The cases were managed conservatively and the scleral folds had resolved within 24 hours. Final VAs of 6/9, 6/9 and 6/18 were consistent with the prior pathologies. Photograph of scleral folds taken 3 hours after surgery (case 1). Ultrasound β-scan of case 1, showing posterior segment collapse with scleral folds. The anterior chamber depth is 4.5 mm. We think that under-perfusion of the posterior segment must have occurred during surgery and that scleral collapse may have been facilitated by orbital pressure on the thin sclera of these myopic eyes. A reverse pupil block would have prevented the free passage of fluid into the posterior segment despite a high infusion pressure into the anterior chamber (Ahfat et al. 2003; Braunstein & Airiani 2003; Misra & Burton 2005). By contrast, misdirection of infusion fluid behind the iris with paradoxical miosis would not exacerbate posterior segment collapse. We propose that reverse pupil block contributed to this phenomenon of scleral collapse by reducing the passage of fluid to the posterior segment during infusion, which prevented the inflation of the posterior segment at the end of the procedure. However, passage of fluid from the posterior segment to the anterior chamber could then still occur when the infusion was stopped, giving a flutter-valve effect. Infusion after lens insertion would then deepen the anterior chamber to give the appearance of normal anatomy while the posterior segment collapse persisted. The phenomenon may have been prevented if the tip of the infusion had been directed between the iris and lens capsule (Cionni et al. 2004; Cheung & Hero 2005). Scleral collapse is a rare event and although the posterior segment appearances were dramatic, the outcome was good without the need for intervention.