Figure 1: Deep stromal localization of air bubbles in a donor eye.
Radial keratotomy as performed by the standard "external" approach requires sophisticated instrumentation. Recently, an ab interno approach1 was described experimentally using a simple blade (Beaver #5910) to complete the procedure and without the need for measurement of corneal thickness. The mean depth of the radial keratotomy incisions using this approach was 58.9% with a mean flattening of 8.20 diopters.1 This "augmented" corneal flattening, despite a relatively shallow radial incision, was attributed to the inverted transverse incision created by the plane of lamellar dissection resulting from the entry of the blade. To increase the depth of the ab interno radial keratotomies (and, therefore, obtain additional corneal flattening) without endothelial penetration, it would be necessary to delineate Descemet's membrane during lamellar dissection by the blade. We found that deep stromal air injection can improve visualization of the deep stroma, allowing deeper radial incisions using the ab interno approach.
Figure 2: Deep stromal spaces corresponding to bubbles from deep intrastromal air injection (hematoxylin and eosin).
Figure 3: Relative depth of the radial incision in 10 eyes without air bubbles. Location of cross section of incision: 0 = central end, 1 = 0.3 mm from centrai end, 2 = 1.55 mm from central end.
Figure 4: Relative depth of the radial incision in eight eyes with deep stromal air bubbles. Locations 0, 1 , and 2 are defined in Figure 3.
MATERIALS AND METHODS
Fresh human donor eyes were harvested and maintained in Neosporin ophthalmic solution at 0° C a few hours after enucleation. The eyes were defrosted at room temperature. The surgical procedure was performed using the Zeiss operating microscope (Thornwood, NY) and immobilizing the eyes in an eye holder. Eighteen human donor eyes underwent two "ab interno" radial keratotomy1 incisions 180° apart. The keratotomy was performed by raising the intraocular pressure to 45 mm Hg as described elsewhere.1 Radial keratotomy was performed using a Beaver blade #5910. The sclera was entered tangentially 1 mm from the limbus up to the edge of the optical zone while staying in the deep stroma.1 The shaft was rotated 90°, allowing the cutting edge to dissect from the deep stroma to the epithelium in a radial fashion. In eight eyes, the radial keratotomy incisions were preceded, along the same meridian, by air injection using a 30-gauge needle. The tip was engaged at the limbus in the deep stroma and advanced up to 3 mm from the limbus. Air was injected in a pulse fashion using a 3-milliliter air filled syringe until the bubbles filled the meridian. Following radial keratotomy, the corneas were excised along with a 2-millimeter scleral rim and fixed in 10% buffered formalin. The corneas were examined histologically to determine the level of the bubbles. To determine the depth of the radial keratotomy incision, the corneas were bisected and the two halves were submitted for paraffin block sectioning and staining with hematoxylin and eosin. The sections were made perpendicular to the plane of the incision. Step sectioning was performed at three levels: the most central part of the incision (level 0), 0.3 mm from level 0 (level 1), and 1.55 mm from level 0 (level 2). The relative depth of one of two incisions (selected at random) from each eye was measured with a Zeiss ocular grid.
Figure 5: Representative cross-section of the radial keratotomy incision in an eye without bubbles.
Figure 6: Representative cross-section of the radial keratotomy incision in an eye with deep stromal bubbles.
The bubbles were located by slit-lamp biomicroscopy (Fig 1) and histologically (Fig 2) in the deep stroma (pre-Descemet region) and less often in the midstroma. The stromal lamellae were smoothly separated by the gas bubbles.
Performance of the "ab interno" radial keratotomy was facilitated by the presence of deep stromal bubbles. The surgeon was able to engage the #5910 Beaver blade in the pre-Descemetic stroma as the cornea was thicker and Descemet's membrane was outlined and "protected" by bubbles. The average (standard deviation) relative depth of the radial incisions in 10 eyes without bubbles (Fig 3) versus 8 eyes with bubbles (Fig 4) was respectively: level 0-58.3% (13.6%) vs 89.3% (4.4%) (student ¿-test t = 5.35; ? < .001); level 1-57.1% (11.8%) vs 81.5% (10.2%) t = 4.40; ? < .001); level 2-57.6% (10.2%) vs 79.1% (13.2%) (t = 3.80; ? < .01); three levels combined- 57.7% (11.5%) vs 83.0% (10.6%) (Figs 5-6).
The localization of bubbles to the pre-Descemet and midstromal regions indicates that these two planes are easier to separate than other planes. Likewise, intracorneal hematomas were found to sequestrate in pre-Descemet and midstromal locations.25 Corneal bubbles from air6 and from perfluoropropane gas used in pneumatic retinopexy (unpublished data) resolve without sequelae within 48 hours. The bubbles probably dissect through the interlamellar proteoglycan bonds7 to form "holes" or "canals."8
Air bubbles facilitate the lamellar dissection by the blade in ab interno radial keratotomy. The author found a marked facilitation of deepilamfellar dissection in six human bank eyes (unpublished data), in a similar way to that reported clinically by Arenas- Archila.6 Enhanced visualization of the deep stromal layers coupled with thickening of the cornea by the bubbles allowed the surgeon to guide the advancing blade into the pre-Descemet space. Using the bubbles as a marker of depth and as a cushion against endothelial penetration allowed the radial keratotomy to be deeper using the ab interno approach. Combining two simple and nonexpensive procedures using a blade and a 30-gauge needle appear promising. Additional experimentation and refinement of the present techniques are underway.
1. Mansour AM, Rowsey JJ, Munn AR III. "Ab interno" radial keratotomy. Refract Corneal Surg. 1991;7:182-185.
2. Muenzler WS. Intracorneal hemorrhage. Am J Ophthalmol. 1973;76:440-443.
3. Searl SS, Croll SD, Boruchoff SA1 Albert DM. Corneal hematoma. Arch Ophthalmol. 1984;102:1647-1649.
4. Ormerod LD, Egan KM. Spontaneous hyphema and corneal hemorrhage as complications of microbial keratitis. Br J Ophthalmol. 1987;71:933-937.
5. Kurz GH. Histopathologic findings in corneas following lens extraction and results of keratoplasty for aphakic bullous keratopathy. Am J Ophthalmol. 1971;72:888-896.
6. Arenas-Archila E. Deep lamellar keratoplasty dissection of host tissue with intrastromal air injection. Cornea. 1984/ 1985;3:217-218.
7. Smolek MK, McCarey BE. Interlamellar adhesive strength in human eyebank corneas. Invest Ophthalmol Vis Sci. 1990;31:1087-1095.
8. Jongebloed WL, Los LI, Worst JgF. Air injection of the corneal stroma: a SEM study. Doc Ophthalmol. 1989; 72:349355.