RESUMO
PURPOSE: To examine the effects of systemic isotretinoin treatment on the eye using several ocular examination parameters. METHODS: We conducted a systemic review for literature published up to June 2021 in both PubMed and Web of Science databases. We included prospective observational or interventional studies evaluating ocular manifestations of isotretinoin in acne patients. The primary outcome measures were anaesthetized and non-anaesthetized Schirmer test, tear break-up time (TBUT), central corneal thickness (CCT), average retinal nerve fibre layer (RNFL) thickness, ganglion cell-inner plexiform layer (GC-IP) thickness, subfoveal choroidal thickness, axial length, ocular surface disease index (OSDI), meibomian gland expression (MGE) and conjunctival stain. The National Institute of Health (NIH) quality assessment tools were used to assess the data quality. The effect size used to analyse the included studies was the weighted mean difference (WMD) and its related confidence intervals (95%CIs). RESULTS: Twenty-one publications involving 1105 eyes of 842 participants met the inclusion criteria. Isotretinoin use was significantly associated with reduction in the scores of anaesthetized Schirmer (WMD = -2.23, 95%CI: -3.28 to -1.18), non-anaesthetized Schirmer (WMD = -3.74, 95%CI: -4.23 to -3.25), TBUT (WMD = -3.47, 95%CI: -5.09 to -1.86), and CCT (WMD= -7.39, 95%CI: -13.91 to -0.88). Isotretinoin use was significantly associated with increase of OSDI (WMD = 18.29, 95%CI: 7.54-29.03), MGE (WMD = 1.02, 95%CI: 0.70-1.33) and conjunctival stain scores (WMD = 0.61, 95%CI: 0.47-0.76). No significant change was noted in RNFL thickness (WMD = -0.64, 95%CI: -1.80 to 0.51); GC-IP thickness (WMD = 0.42, 95%CI: -1.08 to 1.92); subfoveal choroidal thickness (WMD = -1.80, 95%CI: -6.69 to 3.09), and axial length (WMD = 0.08, 95%CI: -0.19 to 0.35). A significant heterogeneity was found between the study estimates in each of anaesthetized Schirmer, TBUT, MGE, OSDI, and conjunctival stain tests. CONCLUSION: Isotretinoin use results in a statistically significant reduction of the central corneal thickness, TBUT, and Schirmer test scores. A statistically significant increase in MGE, OSDI and conjunctival stain scores was found. No statistically significant change of average RNFL, GC-IP thickness, subfoveal choroidal thickness, or axial length was observed. Further well-designed studies should evaluate the long-term effect of isotretinoin on the eye and reach a firmer conclusion.