Radial Optic Neurotomy Using Nasal and Temporal Approach Incisions:  Histopathologic Study in Human Cadaver Eyes FREE

Transvitreal radial optic neurotomy (RON) has recently been described as a surgical management for central retinal vein occlusion,^1- 7 combined cilioretinal artery and central retinal vein occlusion,⁸ and nonarteritic anterior ischemic optic neuropathy,⁹ as well as for the treatment of acute functional impairment associated with optic nerve (ON) drusen.¹⁰ The rationale of RON in central retinal vein occlusion is to alleviate the compartment syndrome that led to occlusion of the central retinal vein, which is housed within the inelastic scleral outlet.¹ RON transvitreally penetrates the scleral ring and lamina cribrosa at the level of the proposed location of the thrombus.¹¹

The mechanism of action of neurotomy is uncertain and has been questioned by some authors.¹² Complications associated with this procedure are visual field defects, vitreous hemorrhage, subretinal hemorrhage, peripapillary retinal detachment, and choroidal neovascularization.^13- 17 There are few studies^2,18- 19 on the histopathologic effects of RON.

It is hypothesized that incision of the full depth of the scleral canal at the ON will allow better blood flow in the central retinal vein in a compartment syndrome model. Because RON is a traumatic procedure performed on delicate and critical structures, safety should be considered before effectiveness. The objectives of the present study were to examine histologically the effects of such incisions on human cadaver eyes and to correlate the nasal and temporal approaches for performing RON on the nasal side of the optic disc using the dominant and nondominant hands in terms of depth and width of ON penetration, vessel wall damage, and globe perforation.

Nine human cadaver eyes were selected with consent from the institutional eye bank for the RON procedure by a right-handed surgeon (M.A.A.) using an unmodified 20-gauge microvitreoretinal (MVR) blade having sharp edges on both sides in a transvitreal approach. Following removal of a sclerocorneal button with limbal margins, 2 sclerotomies were created using the MVR blade for the introduction of a fiberoptic light probe and a 20-gauge MVR blade. An operating microscope was used for visualization and illumination. A radial incision into the nasal margin of the ON was performed using the MVR blade without removal of the vitreous. A nasal approach was used in 4 left eyes (using the right hand) and in 2 right eyes (using the left hand) (Figure 1A and B), and a temporal approach (Figure 2) was used in 3 right eyes (using the right hand).

All tissue samples were embedded in paraffin, and 5-μm sections were obtained in various orientations (1 axial, 1 sagittal, and 7 coronal). Microscopic examination of permanent hematoxylin-eosin slides was used to assess for scleral canal incision and to determine whether damage to ON vasculature occurred. The mean ON area of decompression was measured using commercially available software (Optimus version 6.5; Media Cybernetics, LP, Silver Spring, Maryland). The depth of ON penetration was measured by counting the number of serial sections comprising the incision and then multiplying by 5 μm per section to determine the total depth of ON penetration (Figure 3). Using the Optimus viewer, the mean ON area of the proximal-most slide containing the incision and the distal-most slide containing the incision was taken to calculate the mean ON area (Figure 4A, B, and C). Every fourth section within the incision was viewed using the Optimus viewer, the areas of intra-ON incision (decompression area) were measured, and the areas were averaged to calculate the mean decompression area. The percentage decompression area was calculated as the ratio of the mean decompression area divided by the mean ON area. All incision slides were reviewed; the slide with the greatest radial ON incision length was determined and was measured linearly to obtain the maximum length of ON penetration. This incision was then measured circumferentially, and this was recorded as the maximum width of ON penetration. Finally, the incision slides were reviewed, and the greatest length of aggregate ON and retinal and scleral incision was determined and was measured linearly to calculate the maximum cut length.

In 7 eyes, coronal sections were of adequate quality for complete analysis. In 2 eyes, the axial and sagittal sections provided information on depth of penetration and potential damage to critical structures (Table). In 8 of 9 eyes, the central retinal vein and artery remained undisturbed. In a single right eye approached from the temporal side using the right hand, the adventitial sheath of the central retinal artery was lacerated, while the lumen, tunica media, and tunica intima were unaffected (Figure 5). The scleral canal was fully incised in all cases, with a mean depth of 555 μm. The mean depth of ON penetration was 725 μm in eyes in which the nasal approach was used and 246.7 μm in eyes in which the temporal approach was used (P =.12). The mean area of ON decompression was 1.12% of the ON area. The nerve fiber layer was minimally disrupted in all eyes. The globe was not ruptured in any eye.

In addition to containing the ON, the scleral outlet is the space through which the central retinal artery and central retinal vein pass into and out of the eye. The scleral outlet contains the lamina cribrosa and is encompassed by the scleral ring. As the ON approaches the eye, it consists of myelinated nerve fibers, central retinal artery, and central retinal vein and has a diameter of 3.0 mm. However, the internal diameter of the optic disc and scleral outlet is 1.5 mm. Therefore, it has been suggested that relaxation of the scleral outlet by RON might be an effective surgical treatment option for scleral outlet compartment syndromes,¹ including central retinal vein occlusion. In view of the potential serious complications of RON, Opremcak et al¹ stressed the importance of a consistent incision; the incision is created in a radial fashion using an MVR blade on the nasal side of the optic disc, approaching the center of the cribriform plate with an insertion depth of just beyond the widest portion of the diamond-shaped tip. Modifications of the MVR blade using a blunt lancet² and a guarded RON knife²⁰ have also been used to produce a safe incision. Although care should be taken to avoid major retinal vessels, it can be difficult to do so, and a case of central retinal artery occlusion has been reported after RON.²¹

Opremcak et al¹ used human cadaveric eyes to investigate the anatomy of the scleral outlet and the best approach to relax the cribriform plate and scleral ring, as well as to determine the location, angle, and depth of penetration using the MVR blade without globe perforation. Measurements for these incisions were not reported. Czajka et al¹⁸ performed histologic examination of the ON after RON in 14 porcine eyes and demonstrated foci of hemorrhage, interstitial edema, reactive gliosis, and rare inflammatory cells with complete axonal nerve fiber loss distal to the neurotomy site at 3 weeks. In a surgical technique study of cadaver human, porcine, and rabbit eyes, Lit et al² demonstrated that puncture of the lamina cribrosa using a specially designed lancet tip having a sharp cutting edge on 1 side and an opposing blunt edge was possible without serious injury to the ON. This procedure differs from the technique and instrumentation used during RON. Tao et al¹⁹ studied the histopathologic findings of normal miniature pig eyes after RON and found localized ON atrophy at the incision site.

The histopathologic examination in our study shows that RON in the human cadaver eye reproducibly results in lysis of the scleral canal at the ON head. Damage to the peripheral wall of a major retinal vessel occurred in 1 eye in which the incision was approached temporally. This approach impairs the surgeon's view of the central retinal vessels (Figure 2). Adverse events such as this can be avoided by always approaching the incision from the nasal aspect of the nerve even if this requires the surgeon to use the nondominant hand (Figure 1). Clear corneal phacoemulsification using the nondominant left hand has been previously shown to be safe and efficacious.²² The procedure of incising the optic disc once at the nasal aspect using the nondominant hand is comparatively brief and should be safe for many retina surgeons. Comfort with the procedure can be enhanced through wet laboratory practice.

A recent histologic study²³ of 111 human globes found a mean ± SD central lamina cribrosa thickness of 378.1 ± 117.8 μm and a mean ± SD scleral thickness at the optic disc border of 276.7 ± 76.1 μm (range, 120-540 μm). Our study demonstrated a greater mean depth of penetration with incisions created from a nasal approach (725 μm) vs a temporal approach (246.7 μm).Therefore, incisions from a nasal approach are more likely to completely incise the scleral ring and peripheral lamina cribrosa than incisions from a temporal approach. This is due to the more oblique angle of penetration with the temporal approach. It is unknown if a complete incision of the scleral ring is required to potentially relax the compartment, but it seems more likely to result in such relaxation than a partial incision. In addition, the obvious practical improvement in visualization would argue for incisions made using a nasal approach.

The histologic findings of RON in human cadaver eyes support the anatomical claims made by originators of the procedure. RON in the human cadaver eye reproducibly results in incision of the scleral canal at the ON head without globe rupture occurring. We cannot conclude from this study if there are secondary changes in the central retinal vein. Damage to major retinal vessels can more reliably be avoided by approaching the RON incision from the nasal aspect of the ON using the dominant or nondominant hand rather than always using the dominant hand.

Correspondence: Michael M. Altaweel, MD, Department of Ophthalmology and Visual Sciences, University of Wisconsin, 2870 University Ave, Ste 206, Madison, WI 53705 (mmaltaweel@wisc.edu).

Submitted for Publication: October 12, 2005; final revision received March 14, 2007; accepted March 21, 2007.

Author Contributions: Dr Altaweel had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Financial Disclosure: None reported.

Disclaimer: Dr Albert, the journal's editor, was not involved in the editorial review or decision to publish this article.