Khurana RN, Patel VR. Paracentral and cecocentral scotomas after pars plana vitrectomy for rhegmatogenous retinal detachment. Am J Ophthalmol 2020;219:163-169.
Pars plana vitrectomy (PPV) is the most commonly employed procedure in the repair of rhegmatogenous retinal detachment (RRD). While preoperative factors, especially the degree of macular detachment, most often determine postoperative visual outcomes of RRD, vision loss may also occur due to intraoperative or postoperative events, such as intraocular bleeding, vascular occlusions, ischemic optic neuropathy, glaucoma, epiretinal membrane and presumed silicone oil toxicity. The featured paper describes an additional rare, but serious, complication of RRD repair by PPV: development of a new paracentral or cecocentral scotoma, separate from any pre-existing scotoma corresponding to the RRD.
The paper consists of a retrospective case series of patients who complained of a new paracentral or cecocentral scotoma following PPV for RRD repair by a single surgeon over a 6-year period. Records of history and examination findings, imaging, and automated perimetry were reviewed. All patients underwent retrobulbar anesthesia with a mixture of 2% lidocaine and 0.75% bupivacaine, followed by either 23- or 25-gauge PPV with valved vitreous cannulas and the Alcon Constellation system set at an infusion pressure of 25 mm Hg. Following fluid-air exchange, all patients had iso-expansile sulfur hexafluoride (SF6) or perfluoropropane (CF3) gas tamponade manually injected into the eye via a 60-mL syringe attached by the infusion tubing to the inferotemporal vitreous port, while a chimney cannula attached to the superotemporal vitreous port allowed venting of air.
Nine of 140 patients (6.4%) undergoing PPV for RRD repair during the 6-year data collection period developed a new scotoma postoperatively. The average age of affected patients was 61 years (range 53-71). All 9 developed a paracentral or cecocentral scotoma involving the inferotemporal visual field and in the right eye only, documented on 10-2 and/or 24-2 visual fields performed with the Humphrey analyzer; in two patients, the scotomas also involved the center of fixation. All 9 scotomas were noticed within the first month postoperatively, and 7 of them were noticed by postoperative week 1. Seven of the 9 patients had macula-on RRD preoperatively and 8 of 9 underwent 23-gauge PPV. No patient had any history of glaucoma or other optic neuropathy, and prior procedures (in 6 patients) included laser retinopexy, pneumatic retinopexy, and (in one patient) a PPV. There were no documented intraocular pressures above 21 mm Hg during the postoperative periods of the affected patients. No retinal whitening was observed on examination, and multimodal imaging including spectral-domain OCT, fluorescein angiography, and fundus autofluorescence was notable only for focal superonasal atrophy of the macular ganglion cell layer, involving the papillomacular bundle and corresponding to the deficits on automated perimetry. All patients had follow-up data through a minimum of 6 months, and for at least some of the patients the scotomas did not resolve, although we are not given details on the extent of improvement, if any.
A similar complication was reported in the 1990s following PPV for macular hole repair, with more peripheral scotomas. This was thought to be due to a high-pressure infusion of dehydrated air during fluid-air exchange, and the incidence of these peripheral scotomas declined following implementation of humidified air and a lower infusion pressure. The authors of the present paper propose a similar mechanism, namely excessive pressure on the superonasal macula from a stream of gas rapidly injected through the inferotemporal vitreous port, with rapid injection of gas facilitated by simultaneous decompression of the eye through a chimney cannula inserted into one of the superior vitreous ports. The authors recommend a slow injection of gas into the eye with or without a more controlled release of intraocular pressure achieved by opening one of the other vitreous ports with toothed forceps, rather than a chimney cannula. This hypothesis is especially plausible when we recall that macular damage (including iatrogenic macular hole) has been described due to rapid injection of dyes into the vitreous cavity for staining of the internal limiting membrane. In any case, the authors’ recommendations are reasonable and can be easily implemented without materially affecting surgical time. Although this study would benefit from a larger sample size and more follow-up data, it describes a unique, serious, likely irreversible, and likely preventable complication of pars plana vitrectomy of which all vitreoretinal surgeons should be aware.
Vlad Matei, MD
Denver Retina Center