High fluence iontophoretic corneal collagen cross-linking: in vivo OCT imaging of riboflavin penetration.

To the Editor: We read with interest the excellent article by Malhotra et al.1 regarding in vivo estimation of ribofl avin penetration using anterior segment optical coherence tomography (OCT). The article evaluates the effect of complete versus grid-like epithelial removal on ribofl avin penetration during collagen cross-linking (CXL) in vivo using hand-held OCT. Twenty eyes of 20 patients were imaged intraoperatively at 30 and 60 minutes after starting the procedure. Results showed a homogeneous hyperrefl ective band extending to a mean depth of 54.2 μm after 30 minutes. In the gridlike removal group, the band was reported uneven in the epi-on areas. We agree with the authors on the use of OCT for in vivo evaluation of ribofl avin’s penetration inside the stroma. Moreover, we agree that grid removal of the epithelium is not able to soak the corneal stroma evenly like the epi-off procedure.2 In our study, we report the in vivo ribofl avin penetration during CXL performed with iontophoresis versus conventional epithelium-off protocol using high-resolution OCT. Six eyes (6 patients) undergoing CXL were measured preoperatively, intraoperatively, and postoperatively using high-resolution OCT. The epithelium was removed completely in the central 9-mm zone in 3 eyes (epi-off group), whereas ribofl avin penetration through intact epithelium was promoted by an iontophoresis device in the remaining 3 eyes (iontophoresis group). The iontophoresis device for corneal application (8 mm in diameter) is placed on the cornea using an annular suction ring (low suction created by a syringe connected on the suction annulus). The device is fi lled with approximately 0.5 mL solution from the open proximal side, until the electrode (stainless steel mesh) is covered (Figure 1A). The device is connected to a constant current generator (I-ON XL, Sooft, Italy) set at 1 mA (the total dose of 5 mA min is monitored by the generator). During the irradiation phase (10mW/cm2 for 9 minutes), a good fl uorescence is clearly detectable (Figure 1B) even after washing with balanced salt solution. The depth of the hyperrefl ective band (representing penetration of ribofl avin) in the anterior corneal stroma was measured. In the conventional epi-off group, after 30 minutes of passive impregnation, a homogeneous hyperrefl ective band without fading effect was measured at mean depth of 80 μm (Figure 1C). In the iontophoresis group, we observed a less homogeneous but deeper hyperrefl ective band with a fading effect extending through the anterior 200 μm of the cornea (Figure 1D). This band was not visible until the end of irradiation time. Intraoperative OCT imaging could be a useful technique to evaluate in vivo penetration of ribofl avin in-