Impact of epiretinal membrane on biometry measurements accuracy in combined phaco-vitrectomy

Abstract

Abstract Background To assess the reproducibility of axial length measurements and the postoperative refractive error in patients undergoing combined phacovitrectomy for idiopathic epiretinal membrane removal. Methods Ten eyes of 10 patients who underwent combined phacovitrectomy were enrolled in this prospective observational case series. The main outcome was the axial length changes on ultrasound biometry. Secondary outcome measures were: changes in central macular thickness (CMT), the correlation between axial length changes on ultrasound biometry and absolute change in central macular thickness, mean predicted refractive error either with optical or ultrasound biometry, absolute refractive error in spherical equivalent (SE), axial length (AL), BCVA and intraocular pressure (IOP). Optical (IOL master 500, Carl Zeiss) and Ultrasound Immersion (Compact Touch, Quantel Medical) axial length together with central macular thickness (OCT – Optical coherence Tomography, Heidelberg) measurements were obtained preoperatively and at 3 months follow up. Intraocular lens power selection was based on Holladay I, SRK-T and Hoffer Q formulas and the preoperative mean predicted refraction calculated either with ultrasound or optical biometry was compared with mean postoperative spherical equivalent measured with automatic refractometry (Nidek, ARK1). A comprehensive assessment including BCVA, Intraocular pressure, refractive status, anterior segment biomicroscopy and fundus examination was carried out at baseline and repeated one day, one and three months postoperatively. Results Mean patient’s age was 70,7 ± 4,83. The mean logMAR BCVA preoperatively was 0,44 ± 0,17 and improved significantly to 0,75 ± 0,16 at 3 months after surgery (p = < 0.05). Ultrasound (US) immersion AL increased (p < 0.05) while CMT decreased significantly (p < 0.05). A correlation was found between US and CMT reduction thus not statistically significant (p > 0.05). Contrary the optical AL decreased postoperatively (p < 0.05). The preoperative macular thickness and the postoperative optical dioptric shift correlated (p = 0.0557) while no correlation was found between the US biometry and the postoperative dioptric shift (p > 0.05). A correlation was also recorded between either the ultrasound (rho = 0,9; p < 0.05) or optical (rho = 0,511; p > 0.05) predicted refraction and the actual refractive outcome as well as between the two techniques. The dioptric shift, was then calculated both for Ultrasound (0,21 ± 0,82 D) and Optical biometry (0,113 ± 0,34 D). Bland-Altman diagram shows the difference between the two shifts was likely to be dependent on the average shift (p < 0.05 Pitman’s test for differences in variance). Conclusions Optical biometry represents the gold standard technique for the AL measurement in combined phacovitrectomy for iERM. The Ultrasound immersion biometry seems to be less accurate and deliver a less predictable refractive outcome, either hyperopic or myopic. This method should be reserved in selected cases and the patient should be always informed about the possibility of a refractive error. Further studies with a larger population and a control group are necessary to expand our understanding in IOL selection in combined surgery and eventually confirm this preliminary results.