- Email: [email protected]
Ten Caveats in Keratorefractive Surgery
Ten Caveats in Keratorefractive Surgery J. JAMES ROWSEY, MD
Abstract: Topographic shifts in the cornea are associated with refractive changes as follows: (1) The normal cornea flattens over any incision. (2) Radial incisions flatten the adjacent cornea and the cornea 90° away. (3) The cornea flattening effect increases as incisions approach the visual axis. (4) The cornea flattens directly over any sutured incision. (5) The cornea flattens adjacent to loose sutures and flattens 180° away, and steepens 90° away. (6) The cornea steepens adjacent to tight sutures and steepens 180° away, and flattens 90° away. (7) The cornea flattens over wedge resection or tucks. (8) The cornea steepens anterior to wedge resections or tucks. (9) Tissue removal produces corneal flattening over the site of tissue removal, whether traumatic or surgically induced. (10) Full-thickness corneal tissue addition produces corneal steepening over the site of the tissue addition and flattens the adjacent cornea. [Key words : corneal sutures, corneal topography, corneoscopy, keratorefractive surgery, relaxing incisions, wedge resections.] Ophthalmology 90:148-
155, 1983
Topographic corneal shifts occur during and following cataract surgery, keratorefractive surgery, and trauma to the eye. Basic similarities in the effect of surgical procedures on the refractive potential of the cornea have been analyzed, and ten basic topographic shifts are observed. A careful analysis of these refractive alterations has allowed us to design surgical approaches to correct spherical refractive errors and astigmatic errors in the cornea. Photographed corneascope Placido disc imagery has been used to document the dioptric power shifts of the cornea. I •2
1. THE NORMAL CORNEA FLAITENS OVER ANY INCISION Any traumatic or surgical incision in the cornea produces flattening of the cornea in the area of the incision. From the Dean A. McGee Eye Institute, Department of Ophthalmology, University of Oklahoma Health Sciences Center, Oklahoma City, Okla· homa. Supported by NEI Grant #5 R01 EY03756·02, an unrestricted grant from Research to Prevent Blindness, the Pfeiffer Foundation, and the private philanthropy of the citizens of Oklahoma. Presented at the Eighty·sixth Annual Meeting of the American Academy of Ophthalmology, Atlanta, Georgia, November 1-6, 1981. Reprint requests to J. James Rowsey, MD, McGee Eye Institute, De· partment of Ophthalmology, University of Oklahoma, 608 Stanton L. Young Boulevard, Oklahoma City, OK 73104.
148
Immediately after corneal incisions an anterior lamellar dehiscence is observed that becomes edematous. The open anterior flat surface of the cornea and intact Descemet's membrane effectively increases the radius of curvature of the cornea, decreasing plus power. To determine the magnitude of this corneal power shift, eyebank eyes were maintained at uniform pressure between 20 and 30 mmHg, and surgical incisions were placed to a depth of 90% of central corneal pachymetry. Photographs were taken of the cornea with a central tattoo mark localizing the center of the proposed optical zone. Two horizontal incisions were placed in these eyebank eyes. Paired incisions traversing the limbus produced the marked corneal curvature changes outlined in Table 1. In all instances, the cornea flattened over the incision site, Ring 3, in the horizontal meridian or the incision meridian (3-H), more than the cornea flattened over the third corneascope ring in the vertical (3-V) meridian 90° away from the incision. Less corneal flattening was noted over the fifth corneascope ring and even less corneal flattening was noted at the ninth ring area. Similar corneal flattening was obtained when incisions did not traverse the limbus. . Limbal incisions for cataract surgery also gape, thereby flattening the adjacent cornea. The effect ofthis superior corneal flattening may not be recognized with central corneal keratometry, but is easily observed by Placido disc analysis of flat rings or markedly disparate rings at the limbus. I This limbal flattening is frequently incompatible with good contact lens fitting, for the contact
0161-6420/83/0200/0148/$1.20 © American Academy of Ophthalmology
ROWSEY • KERA TOREFRACTIVE SURGERY
Table 1. Radial Keratotomy Eyebank Data-Effect of Two Horizontal Incisions: Curvature Change in Diopters 3V
3H Optical Zone 3 mm-limbus
Eye # 5 8 9 15
Mean
Range low high
Ring number
5H
5V
9H
9V
-8.25 -6.88 -4.62 -8.12
-2.88 -5.50 -0.75 -1.12
-6.00 -5.00 -1.75 -4.50
-0.88 -4.62 -1.88 +0.37
-0.50 -0.62 +0.88 -1.62
+0.25 -0.88 -2.88
-6.97 4.63 8.25
-2.56 0.75 5.50
-4.31 1.75 6.00
-1.75 0.37 4.62
-0.47 0.50 1.62
-1.17 0.25 2.88
3-10 mm
21 22 23
-4.25 -4.75 -11.50
-2.00 -2.00 -2.75
-4.25 -5.25 -9.25
-2.00 -2.25 -0.50
-1.25 -1.25 -4.00
0 -1.75 -.50
3-11.1 mm
24
-4.25
-2.75
-3.75
-4.25
-2.25
-1.75
-6.19 4.25 11.50
-2.38 2.00 2.75
-5.62 3.75 9.25
-2.25 0.50 4.25
-2.18 1.25 4.00
-1.00 0.50 1.75
Mean Range low high
Radial incisions in eyebank eyes produce marked corneal flattening along the axis of the incisions which were placed in the horizontal meridian by convention. The vertical meridian also demonstrates corneal flattening
of approximately half the magnitude demonstrated in the horizontal merid ian centrally.
lens rides inferiorly away from the flat area of the incision (Fig 1). Unfortunately, traumatic incisions around the visual axis are often associated with marked corneal flattening centrally and unequal flattening in the peripheral cornea, producing irregular astigmatism. Controlled flattening of the cornea over an incision provides the potential for myopic correction in radial keratotomy or astigmatism surgery, approximating the power shifts outlined in Table 1. Figure 2 demonstrates the increased flattening seen along incisions and the disparate flattening along both sides of the same incision due to edema. Inflammatory thinning of the cornea may be associated with marked topographic shifts and induced astigma-
tism. The thinned cornea is no longer normal and, therefore, does not necessarily respond to incisions by flattening. The dioptric shifts observed in Table 1 cannot be expected in thin corneas.
WOUND GAPE
FLAT
V\.~@lqn
Fig 1. Caveat I. Schematic of contact lens fitting over the flat superior cornea. The central radius of curvature of the cornea is compatible with good contact lens fitting, but the flat superior cornea allows the contact lens to ride inferiorly off the visual axis.
2. RADIAL CORNEAL INCISIONS FLATfEN THE ADJACENT CORNEA AND THE CORNEA 90° AWAY As incisions approach within 1.5 mm of the visual axis (3-mm optical zone), the cornea is flattened approximately 6 diopters over the incisions themselves. Ninety degrees away, approximately half this magnitude of flattening is observed in the eyebank eye in vitro and clinically in vivo (Fig 3). Astigmatic error correction in vivo by placing radial incisions around the carefully documented visual axis is a viable consideration. Since approximately 6 diopters of flattening are observed along a corneal incision and 3 diopters may be accomplished at 90°, a patient with a refraction of -6.00 +3.00 X any axis could anticipate a large measure of correction when relaxing incisions are placed along the axis of the plus power cylinder. Although circumferential incisions similarly flatten the cornea, simultaneous radial and circumferential incisions should be avoided. Combined incision techniques are associated with poor centripetal migration of epithelial cells from the limbal palisades of Vogt, and subsequent epithelial erosions have been observed in nonhuman primates and in patients. 149
OPHTHALMOLOGY • FEBRUARY 1983 • VOLUME 90 • NUMBER 2
3. THE CORNEA FLATTENING EFFECT INCREASES AS INCISIONS APPROACH THE VISUAL AXIS As incisions traverse the visual axis, maximum flattening of the cornea occurs. Although flattening occurs over the entire incision (Caveat I), flattening is greatest in the central corneal area of the incision and diminishes rapidly in the midperipheral and peripheral cornea (Table 2). A 3-mm central optical zone produces marked central cornea flattening of approximately 10.5 diopters in eyebank eyes, and over 5 diopters of permanent corneal flattening one year after radial keratotomy in patients. 3 Although 10 diopters of central corneal flattening on corneoscopyat Ring 3 are observed, only 7.5 diopters of midperipheral corneal flattening and 2.5 diopters of Ring 9 flattening are observed when a 3-mm optical zone is used. Radial incisions in the cornea with a clear, 6-mm optical zone produce minimal central corneal flattening, but over 4 diopters of flattening over the incisions themselves. A large 9-mm optical zone radial keratotomy is associated with one diopter of flattening over the incisions, but steepening in the center of the cornea instead of flattening. Figure 4 demonstrates the additional effect of radial keratotomy of extending the incisions closer to the visual axis. Corneal flattening is, therefore, not equal along radial incisions and the effect upon visual axis flattening is predicted by the proximity of the incisions to the central visual axis. Marked central flattening diminishes rapidly in the midperiphery ofthe cornea after radial keratotomy (Figs 2, 4). A mid peripheral "knee" is thereby produced in the cornea after a radial incision. The disparate flattening of the cornea after radial keratotomy produces optical aberrations that may be disconcerting to the surgeon and the patient.
Two of the main difficulties after radial keratotomy are glare and fluctuating vision. The optics of the midperipheral corneal "knee" may act as a plus lens system, producing glare and induced myopia with pupil dilation at the end of the day4 (Fig 5). We have noted a bull's eye retinoscopy reflex after radial keratotomy analogous to keratoconus optical aberrations. In keratoconus, the peripheral ring of the retinoscopy image is neutralized, and the central "bull's eye" requires additional plus power to neutralize fully the keratoconus reflex. The peripheral corneal-induced power provides an "against" motion when the center of the cornea is finally neutralized. After radial keratotomy, however, the central cornea is flatter than the peripheral cornea and is neutralized first with addition of plus spheres due to its lower plus power. The peripheral knee requires additional plus power for subsequent neutralization, and the center of the cornea demonstrates "against motion" when the peripheral "knee" is neutralized. These optical aberrations in radial keratotomy are accentuated with pupil dilation and patients may experience a precipitous decrease in vision in the evening as their mobile pupil dilates.
4. THE CORNEA FLATTENS DIRECTLY OVER ANY SUTURED INCISION Regardless of the method of suturing a corneal incision, corneal flattening occurs. Loose sutures in a limbal corneal incision or in clear stromal incisions allow the normal flattening of the edematous incised cornea to be observed. Tight sutures might be expected to produce corneal steepening if flattening of the cornea occurs with loose sutures. However, the compressive effect directly over tight sutures prevents wound gape, and marked
) Fig 2. Top left, Caveat I. Corneoscopy 24 hours following radial keratotomy demonstrates radial incisions with marked flattening in the center of the cornea and partial wound healing. Flattening occurs along the incisions themselves. The longest chord from the center of the photograph to the ring image is observed along the incision (arrow). This 37 diopter cornea demonstrates 9 diopters of flattening. As in map topography, those rings that are markedly disparate or separated from the center apex of the cornea have the longest radius of curvature, and represent the flattest cornea in this area. Minute alterations in corneal topography may therefore be analyzed even along the two sides of an incision. The peripheral rings at the 3 o'clock position (arrow) appear to be broken by the incision and the superior 6th and 7th rings are closer to the center or apex of the cornea than the lower half of the same ring. The lower portion of the cornea is therefore flatter than the superior half of the incision. All of the incision positions, however, are flatter than the preoperative intervening cornea. The faceplate presents dioptric powers of the cornea in the respective rings and corneal positions. The average, or mean corneal power is presented for rings 3-9. We note that the central cornea has flattened to 25-35 diopters after radial keratotomy, but the peripheral cornea remains steeper at 39-40 diopters. Fig 3. Top right, Caveat 2. Following astigmatism surgery with relaxing incisions vertically, a patient demonstrates marked flattening of the cornea in the vertical meridian and an alteration of refractive power from -3.00 +6.00 X 88 to +3.00 +3.00 X 90. Note the marked vertical flattening of the central corneal rings in the 12 to 6 o'clock, or the 90° meridian. This is manifested by a vertical eclipse in the central three rings (arrow). Less flattening is observed in the periphery of the cornea. The cornea flattens in both meridians with radial incisions, however, simultaneously reducing both the astigmatism and the myopia in this patient. The corneascope computer-generated faceplate demonstrates reasonably symmetrical central corneal flattening and a short chord in the periphery in the vertical meridian only, representing residual astigmatism in the peripheral cornea. Fig 4. Bottom, Caveat 3. Apatient who has undergone radial keratotomy with a 5.0 -mm optical zone has sustained I diopter of refractive change. The central cornea flattens from 47 to 45 diopters, and the peripheral cornea flattens from 46 to 45 diopters. When the incisions are extended subsequently to a 3-mm optical zone, the patient develops 3 diopters of refractive change and the central cornea flattens from 46 to 42 diopters. It is apparent that as incisions approach the visual axis, corneal flattening increases. The wide separation of the central rings demonstrates that corneal flattening diminishes (long arrow) as the peripheral cornea is approached, producing a geometrical and optical corneal "knee" at the transition point between the central and peripheral cornea (short arrow).
150
OPHTHALMOLOGY •
FEBRUARY 1983 •
VOLUME 90 •
NUMBER 2
Table 2. Radial Keratotomy Eyebank Data-Effect of 16 Radial Incisions: Curvature Change in Diopters
3H Optical Zone
3 mm-limbus
Mean Range
low high
Eye #
12 15 11 15 9 10 17 3 8 16 18 24 28
3V Ring Number
5H
5V
9H
9V -1.37 -3.62
-7.50 -7.50 -8.75 -13.00 -11.25 -10.75 -11.75 -12.75 -12.12 -11.75 -14.00 -13.75
-8.00 -8.88 -8.25 -5.62 -9.62 -10.00 -9.50 -10.88 -11.62 -12.25 -12.75 -13.25
-2.88 -6.50 -6.75 -5.50 -5.75 -6.88 -7.25 -8.75 -7.50 -7.75 -9.50 -10.00 -11.00
-8.00 -6.50 -5.25 -5.00 -7.88 -7.50 -8.50 -7.50 -9.88 -9.50 -11.00 -9.50
-0.88 -2.62 -4.00 -3.75 +0.50 +0.50 -3.50 -4.25 -1.37 -2.25 -2.50 -4.75 -2.50
-2.50 -4.75 -3.50
-11.24 7.50 13.75
-10.05 5.62 13.25
-7.38 2.88 11.00
-8.00 5.00 11.00
-2.41 0.50 4.75
-2.90 1.37 4.75
When 16 radial incisions are placed in eyebank eyes, symmetrical central corneal flattening occurs of approximately 10.5 diopters. Less
flattening of the cornea is observed (Figs 8A, B). Flattening of the cornea is manifested by ring separation or broadening, indicating a longer (flatter) radius of curvature. Mattress sutures steepen the cornea within the suture area but also flatten the cornea directly under the penetration site of the suture into the corneal stroma.
5. THE LIMBAL CORNEA FLATTENS ADJACENT TO LOOSE SUTURES
-2.12 -3.88 -1.75 -1.75 -4.50 -2.12
midperipheral flattening is observed over the fifth and ninth rings respectively.
6. THE LIMBAL CORNEA STEEPENS ADJACENT TO TIGHT SUTURES (A) AND STEEPENS 180 0 AWAY, (B) AND FLATIENS 90 0 AWAY
Tight suturing of the limbus after cataract surgery may allow for control of preoperative corneal astigma-
(A) AND FLATIENS 180 0 AWAY, (B) AND STEEPENS 90 0 AWAY
The effect of corneal flattening may be advantageous or a disaster after cataract surgery, depending on the preoperative refractive error of the patient. Flattening of the limbal corneal incision with loose sutures (Caveat 1) is accompanied by elongation of the entire arc length of the cornea from the 12 to the 6 o'clock limbus, and elongation of the corneal chord or diameter (Fig 6). Flattening of the cornea in the meridian of loose sutures is routinely associated with concomitant steepening of the cornea 90° away, or in the horizontal meridian after cataract surgery. The normal 43 diopter spherical cornea may demonstrate wound dehiscence with a 38 diopter vertical meridian, (too flat) and a 48 diopter horizontal cornea (too steep). Relaxing incisions along the horizontal meridian would be inappropriate for superior wound gape. We have found it auspicious to reopen cataract incisions suffering from wound gape and resuturing the incisions with nonabsorbable sutures to compress the wound and add plus power to the involved cornea (Caveat 6) when necessary. 152
..
:
r\c/r='===:::~
r
\1
r
;
Fig 5. Caveat 3. Development of midperipheral knee after radial keratotomy is related to flattening of central cornea with less flattening of the peripheral cornea.
ROWSEY •
KERATOREFRACTIVE SURGERY
7. THE CORNEA FLATTENS OVER WEDGE RESECTION OR TUCK
AGAINST THE RULE
"....... I1>I'IU
Fig 6. Caveat 5. Diagram of wound gape with vertical arc (solid curve) and chord (dotted line) elongation following cataract surgery and horizontal shortening of the corneal radius of curvature, producing against-the-rule astigmatism.
tism. Although the cornea flattens directly over any suture, the posterior vector force of tight sutures is compensated by a plus power short radius of curvature effect anterior (central) to the sutures themselves (Fig 7). The balloon corneal model of Troutman5 demonstrates this caveat well. Tight sutures at the limbus produce a short radius of curvature adjacent to the opposite limbus and a long radius of curvature 90° away. Many authors6- 8 feel that slight with-the-rule astigmatism is desirable in the one- to four-week postoperative period following cataract surgery. An immediate wound compression force producing 1.5 to 2.5 diopters of with-the-rule astigmatism resolves following limbal incision suturing, producing an anastigmatic cornea. If sutures traverse the visual axis, no cornea remains centrally for subsequent steepening. Therefore, the total effect on corneal power is due to the corneal flattening itself and no plus power can be induced by tight or loose suturing.
r ..... \ r
~
............
Fig 7. Caveat 6. Following cataract surgery, tight sutures produce wound compression and vertical corneal astigmatism. Although the cornea flattens and lengthens the radius of curvature (r) directly under the suture (arrow), it steepens anterior to the suture producing withthe-rule astigmatism or a short radius of corneal curvature (r) centrally.
Wedge resections may produce marked dioptric power shifts in the cornea. The cornea flattens over a wedge resection more than tight suture contraction may allow. The wider or larger the wedge resection, 8 the larger the degree of flattening that occurs over the wedge resection itself. Wedge resections often involve one full quadrant of the cornea to produce symmetry of plus power anterior to the wedge. Both wedge resection and corneal tucks have diminished effect closer to the limbus and an increased effect as the visual axis is approached. The flattening effect of tissue removal may be more permanent with wedge resection than with corneal tucking procedures. If tissue has filled the interstices of a cataract incision wound gape after one to two years of healing, a wedge resection in the original incision with removal of the proliferative tissue will allow the addition of plus power to the cornea by flattening the limbal area. Wedge resections or tucks may be used in a corneal transplant donor-recipient interface if 10 to 12 diopters of correction are required (wedge), or 5 to 10 diopters of power are needed (tuck).
8. THE CORNEA STEEPENS ANTERIOR TO WEDGE RESECTIONS OR TUCKS The surgeon's goal is plus power addition in the meridian of a wedge resection. If the cornea is flatter than 43 diopters and significant astigmatism is present, the surgeon may anticipate adding plus power in the axis of the cornea where a tuck or wedge resection is accomplished. No tissue is removed during the tuck procedure, but a circumferential incision is placed near the limbus or in clear cornea. The incision is closed by mattress sutures (Fig. 9). Inelastic prolene sutures may cut through the corneal bed, and corneal power may therefore dissipate following a tuck. Large plus power dioptric corrections are possible with wedge resections and less regression may be observed due to permanent structural alterations.
9. TISSUE REMOVAL PRODUCES CORNEAL FLATTENING OVER THE SITE OF TISSUE REMOVAL, WHETHER TRAUMATIC OR SURGICALLY INDUCED Self-healing lacerations of the cornea are frequently not repaired and may be missed by both the patient and the physician. However, any penetrating corneal laceration is associated with flattening and decreased corneal power (Caveat 1). If stromal tissue is extracted by trauma or debrided at the time of surgical repair, the subsequent tight suture required to seal the laceration is routinely associated with marked flattening of the cornea. If tight 153
OPHTHALMOLOGY • FEBRUARY 1983 • VOLUME 90 • NUMBER 2
Fig 8. Caveat 4. A corneal laceration that has been meticulously sutured produces corneal flattening whether the sutures are tight or loose. A horizontal traumatic incision of the cornea has been closed carefully and demonstrates the flattening directly over the incision itself (arrows) by both widening of ring separations and broadening of the rings themselves.
Fig 9. Left, the vector of a compression of a suture is demonstrated as a short chord "V" configuration under the suture itself. A patient who has relaxing incisions and a single compression tuck suture at 4:30 o'clock demonstrates persistence of the symmetrizing effect of the suture " V" in this area (arrow). This vector of compression may be relieved by suture removal at any time in the postoperative period and the induced astigmatism will diminish. If wound compression is not observed, no diminution of corneal astigmatism can be anticipated by suture removal. Fig 10. RighI, Caveats 9 and 10. Corneascope photograph of patient who has undergone tectonic patch graft at 8:00 (arrow) for corneal astigmatism due to traumatic tissue loss. Tissue addition to the cornea steepens the area of tissue addition and flattens the visual axis, thereby reducing corneal astigmatism.
sutures are necessary in the visual axis, then decreased corneal power is realized after surgery. If, however, tight sutures are required in the midperipheral or peripheral cornea, marked flattening in the area of repair is associated with concomitant steepening of the cornea centrally in the visual axis (Caveat 6). Thus, repair of cor154
neal lacerations is similar refractively to an uncontrolled wedge resection of the cornea (Caveats 7 and 8). Suture removal early in the postoperative period may be associated with release of the flattening effect and diminished astigmatism while risking wound dehiscence and leak. In the area of tissue loss, early suture removal does
ROWSEY •
KERATOREFRACTIVE SURGERY
not replace the tissue loss and a flat area of the cornea may persist similar to postinflammatory thinning of the cornea.
10. FULL-THICKNESS CORNEAL TISSUE ADDITION PRODUCES CORNEAL STEEPENING OVER THE SITE OF THE TISSUE ADDITION AND FLATTENS THE ADJACENT CORNEA Corneal stromal incisions with tissue addition corrects the flattening effect observed with traumatic tissue removal. The flat cornea represents a radius of curvature that is excessively long, due to compression. The convex surface of additional corneal tissue reduces the preoperative corneal curvature and corrects both the tissue site and simultaneously flattens the visual axis (Fig 10), essentially reversing the effect of the traumatic wedge resection. Tissue addition to a traumatically flattened area of the cornea may thereby permit the patient to return to contact lens use if necessary.
SUMMARY We have observed ten basic caveats of topographic and keratorefractive change following surgical or traumatic incision in the cornea. Placido disc imagery or
corneoscope photography with subsequent careful review of the dioptric power shifts have allowed documentation of the refractive alterations in these procedures. Planned reconstructive procedures to reduce myopia, hyperopia, or astigmatism take into account these basic caveats. The appropriate removal of sutures following cataract or penetrating keratoplasty surgery is facilitated by an understanding of these ten caveats.
REFERENCES 1. Rowsey JJ, Reynolds AE, Brown R. Corneal Topography: corneascope. Arch Ophthalmol1981; 99:1093-100. 2. Doss JD, Hutson RL, Rowsey JJ, Brown DR. Method for calculation of corneal profile and power distribution. Arch Ophthalmol 1981; 99:1261-5. 3. Rowsey JJ, Balyeat HD. Prelirninary results and complications of radial keratotomy. Am J Ophthalmol 1982; 93:437-55. 4. Dahlstrom R. Personal cornrnunication. 5. Troutman RC. Microsurgery of the Anterior Segment of the Eye, Vol. II. The Cornea: Optics and Surgery. St. Louis: CV Mosby, 1977; 26873. 6. Jaffe NS. Cataract Surgery and its Cornplications. St Louis: CV Mosby, 1976; 83-98. 7. Jaffe NS, Clayrnan HM. The pathophysiology of corneal astigmatisrn after cataract extraction. Trans Am Acad Ophthalmol Otolaryngol 1975; 79:615-30. 8. Barner SS. Surgical induction of corneal astigmatism; An experirnental study. Albrecht von Graefes Arch Klin Exp OphthalmoI1977; 201 :213-20.
155
Abstract: Topographic shifts in the cornea are associated with refractive changes as follows: (1) The normal cornea flattens over any incision. (2) Radial incisions flatten the adjacent cornea and the cornea 90° away. (3) The cornea flattening effect increases as incisions approach the visual axis. (4) The cornea flattens directly over any sutured incision. (5) The cornea flattens adjacent to loose sutures and flattens 180° away, and steepens 90° away. (6) The cornea steepens adjacent to tight sutures and steepens 180° away, and flattens 90° away. (7) The cornea flattens over wedge resection or tucks. (8) The cornea steepens anterior to wedge resections or tucks. (9) Tissue removal produces corneal flattening over the site of tissue removal, whether traumatic or surgically induced. (10) Full-thickness corneal tissue addition produces corneal steepening over the site of the tissue addition and flattens the adjacent cornea. [Key words : corneal sutures, corneal topography, corneoscopy, keratorefractive surgery, relaxing incisions, wedge resections.] Ophthalmology 90:148-
155, 1983
Topographic corneal shifts occur during and following cataract surgery, keratorefractive surgery, and trauma to the eye. Basic similarities in the effect of surgical procedures on the refractive potential of the cornea have been analyzed, and ten basic topographic shifts are observed. A careful analysis of these refractive alterations has allowed us to design surgical approaches to correct spherical refractive errors and astigmatic errors in the cornea. Photographed corneascope Placido disc imagery has been used to document the dioptric power shifts of the cornea. I •2
1. THE NORMAL CORNEA FLAITENS OVER ANY INCISION Any traumatic or surgical incision in the cornea produces flattening of the cornea in the area of the incision. From the Dean A. McGee Eye Institute, Department of Ophthalmology, University of Oklahoma Health Sciences Center, Oklahoma City, Okla· homa. Supported by NEI Grant #5 R01 EY03756·02, an unrestricted grant from Research to Prevent Blindness, the Pfeiffer Foundation, and the private philanthropy of the citizens of Oklahoma. Presented at the Eighty·sixth Annual Meeting of the American Academy of Ophthalmology, Atlanta, Georgia, November 1-6, 1981. Reprint requests to J. James Rowsey, MD, McGee Eye Institute, De· partment of Ophthalmology, University of Oklahoma, 608 Stanton L. Young Boulevard, Oklahoma City, OK 73104.
148
Immediately after corneal incisions an anterior lamellar dehiscence is observed that becomes edematous. The open anterior flat surface of the cornea and intact Descemet's membrane effectively increases the radius of curvature of the cornea, decreasing plus power. To determine the magnitude of this corneal power shift, eyebank eyes were maintained at uniform pressure between 20 and 30 mmHg, and surgical incisions were placed to a depth of 90% of central corneal pachymetry. Photographs were taken of the cornea with a central tattoo mark localizing the center of the proposed optical zone. Two horizontal incisions were placed in these eyebank eyes. Paired incisions traversing the limbus produced the marked corneal curvature changes outlined in Table 1. In all instances, the cornea flattened over the incision site, Ring 3, in the horizontal meridian or the incision meridian (3-H), more than the cornea flattened over the third corneascope ring in the vertical (3-V) meridian 90° away from the incision. Less corneal flattening was noted over the fifth corneascope ring and even less corneal flattening was noted at the ninth ring area. Similar corneal flattening was obtained when incisions did not traverse the limbus. . Limbal incisions for cataract surgery also gape, thereby flattening the adjacent cornea. The effect ofthis superior corneal flattening may not be recognized with central corneal keratometry, but is easily observed by Placido disc analysis of flat rings or markedly disparate rings at the limbus. I This limbal flattening is frequently incompatible with good contact lens fitting, for the contact
0161-6420/83/0200/0148/$1.20 © American Academy of Ophthalmology
ROWSEY • KERA TOREFRACTIVE SURGERY
Table 1. Radial Keratotomy Eyebank Data-Effect of Two Horizontal Incisions: Curvature Change in Diopters 3V
3H Optical Zone 3 mm-limbus
Eye # 5 8 9 15
Mean
Range low high
Ring number
5H
5V
9H
9V
-8.25 -6.88 -4.62 -8.12
-2.88 -5.50 -0.75 -1.12
-6.00 -5.00 -1.75 -4.50
-0.88 -4.62 -1.88 +0.37
-0.50 -0.62 +0.88 -1.62
+0.25 -0.88 -2.88
-6.97 4.63 8.25
-2.56 0.75 5.50
-4.31 1.75 6.00
-1.75 0.37 4.62
-0.47 0.50 1.62
-1.17 0.25 2.88
3-10 mm
21 22 23
-4.25 -4.75 -11.50
-2.00 -2.00 -2.75
-4.25 -5.25 -9.25
-2.00 -2.25 -0.50
-1.25 -1.25 -4.00
0 -1.75 -.50
3-11.1 mm
24
-4.25
-2.75
-3.75
-4.25
-2.25
-1.75
-6.19 4.25 11.50
-2.38 2.00 2.75
-5.62 3.75 9.25
-2.25 0.50 4.25
-2.18 1.25 4.00
-1.00 0.50 1.75
Mean Range low high
Radial incisions in eyebank eyes produce marked corneal flattening along the axis of the incisions which were placed in the horizontal meridian by convention. The vertical meridian also demonstrates corneal flattening
of approximately half the magnitude demonstrated in the horizontal merid ian centrally.
lens rides inferiorly away from the flat area of the incision (Fig 1). Unfortunately, traumatic incisions around the visual axis are often associated with marked corneal flattening centrally and unequal flattening in the peripheral cornea, producing irregular astigmatism. Controlled flattening of the cornea over an incision provides the potential for myopic correction in radial keratotomy or astigmatism surgery, approximating the power shifts outlined in Table 1. Figure 2 demonstrates the increased flattening seen along incisions and the disparate flattening along both sides of the same incision due to edema. Inflammatory thinning of the cornea may be associated with marked topographic shifts and induced astigma-
tism. The thinned cornea is no longer normal and, therefore, does not necessarily respond to incisions by flattening. The dioptric shifts observed in Table 1 cannot be expected in thin corneas.
WOUND GAPE
FLAT
V\.~@lqn
Fig 1. Caveat I. Schematic of contact lens fitting over the flat superior cornea. The central radius of curvature of the cornea is compatible with good contact lens fitting, but the flat superior cornea allows the contact lens to ride inferiorly off the visual axis.
2. RADIAL CORNEAL INCISIONS FLATfEN THE ADJACENT CORNEA AND THE CORNEA 90° AWAY As incisions approach within 1.5 mm of the visual axis (3-mm optical zone), the cornea is flattened approximately 6 diopters over the incisions themselves. Ninety degrees away, approximately half this magnitude of flattening is observed in the eyebank eye in vitro and clinically in vivo (Fig 3). Astigmatic error correction in vivo by placing radial incisions around the carefully documented visual axis is a viable consideration. Since approximately 6 diopters of flattening are observed along a corneal incision and 3 diopters may be accomplished at 90°, a patient with a refraction of -6.00 +3.00 X any axis could anticipate a large measure of correction when relaxing incisions are placed along the axis of the plus power cylinder. Although circumferential incisions similarly flatten the cornea, simultaneous radial and circumferential incisions should be avoided. Combined incision techniques are associated with poor centripetal migration of epithelial cells from the limbal palisades of Vogt, and subsequent epithelial erosions have been observed in nonhuman primates and in patients. 149
OPHTHALMOLOGY • FEBRUARY 1983 • VOLUME 90 • NUMBER 2
3. THE CORNEA FLATTENING EFFECT INCREASES AS INCISIONS APPROACH THE VISUAL AXIS As incisions traverse the visual axis, maximum flattening of the cornea occurs. Although flattening occurs over the entire incision (Caveat I), flattening is greatest in the central corneal area of the incision and diminishes rapidly in the midperipheral and peripheral cornea (Table 2). A 3-mm central optical zone produces marked central cornea flattening of approximately 10.5 diopters in eyebank eyes, and over 5 diopters of permanent corneal flattening one year after radial keratotomy in patients. 3 Although 10 diopters of central corneal flattening on corneoscopyat Ring 3 are observed, only 7.5 diopters of midperipheral corneal flattening and 2.5 diopters of Ring 9 flattening are observed when a 3-mm optical zone is used. Radial incisions in the cornea with a clear, 6-mm optical zone produce minimal central corneal flattening, but over 4 diopters of flattening over the incisions themselves. A large 9-mm optical zone radial keratotomy is associated with one diopter of flattening over the incisions, but steepening in the center of the cornea instead of flattening. Figure 4 demonstrates the additional effect of radial keratotomy of extending the incisions closer to the visual axis. Corneal flattening is, therefore, not equal along radial incisions and the effect upon visual axis flattening is predicted by the proximity of the incisions to the central visual axis. Marked central flattening diminishes rapidly in the midperiphery ofthe cornea after radial keratotomy (Figs 2, 4). A mid peripheral "knee" is thereby produced in the cornea after a radial incision. The disparate flattening of the cornea after radial keratotomy produces optical aberrations that may be disconcerting to the surgeon and the patient.
Two of the main difficulties after radial keratotomy are glare and fluctuating vision. The optics of the midperipheral corneal "knee" may act as a plus lens system, producing glare and induced myopia with pupil dilation at the end of the day4 (Fig 5). We have noted a bull's eye retinoscopy reflex after radial keratotomy analogous to keratoconus optical aberrations. In keratoconus, the peripheral ring of the retinoscopy image is neutralized, and the central "bull's eye" requires additional plus power to neutralize fully the keratoconus reflex. The peripheral corneal-induced power provides an "against" motion when the center of the cornea is finally neutralized. After radial keratotomy, however, the central cornea is flatter than the peripheral cornea and is neutralized first with addition of plus spheres due to its lower plus power. The peripheral knee requires additional plus power for subsequent neutralization, and the center of the cornea demonstrates "against motion" when the peripheral "knee" is neutralized. These optical aberrations in radial keratotomy are accentuated with pupil dilation and patients may experience a precipitous decrease in vision in the evening as their mobile pupil dilates.
4. THE CORNEA FLATTENS DIRECTLY OVER ANY SUTURED INCISION Regardless of the method of suturing a corneal incision, corneal flattening occurs. Loose sutures in a limbal corneal incision or in clear stromal incisions allow the normal flattening of the edematous incised cornea to be observed. Tight sutures might be expected to produce corneal steepening if flattening of the cornea occurs with loose sutures. However, the compressive effect directly over tight sutures prevents wound gape, and marked
) Fig 2. Top left, Caveat I. Corneoscopy 24 hours following radial keratotomy demonstrates radial incisions with marked flattening in the center of the cornea and partial wound healing. Flattening occurs along the incisions themselves. The longest chord from the center of the photograph to the ring image is observed along the incision (arrow). This 37 diopter cornea demonstrates 9 diopters of flattening. As in map topography, those rings that are markedly disparate or separated from the center apex of the cornea have the longest radius of curvature, and represent the flattest cornea in this area. Minute alterations in corneal topography may therefore be analyzed even along the two sides of an incision. The peripheral rings at the 3 o'clock position (arrow) appear to be broken by the incision and the superior 6th and 7th rings are closer to the center or apex of the cornea than the lower half of the same ring. The lower portion of the cornea is therefore flatter than the superior half of the incision. All of the incision positions, however, are flatter than the preoperative intervening cornea. The faceplate presents dioptric powers of the cornea in the respective rings and corneal positions. The average, or mean corneal power is presented for rings 3-9. We note that the central cornea has flattened to 25-35 diopters after radial keratotomy, but the peripheral cornea remains steeper at 39-40 diopters. Fig 3. Top right, Caveat 2. Following astigmatism surgery with relaxing incisions vertically, a patient demonstrates marked flattening of the cornea in the vertical meridian and an alteration of refractive power from -3.00 +6.00 X 88 to +3.00 +3.00 X 90. Note the marked vertical flattening of the central corneal rings in the 12 to 6 o'clock, or the 90° meridian. This is manifested by a vertical eclipse in the central three rings (arrow). Less flattening is observed in the periphery of the cornea. The cornea flattens in both meridians with radial incisions, however, simultaneously reducing both the astigmatism and the myopia in this patient. The corneascope computer-generated faceplate demonstrates reasonably symmetrical central corneal flattening and a short chord in the periphery in the vertical meridian only, representing residual astigmatism in the peripheral cornea. Fig 4. Bottom, Caveat 3. Apatient who has undergone radial keratotomy with a 5.0 -mm optical zone has sustained I diopter of refractive change. The central cornea flattens from 47 to 45 diopters, and the peripheral cornea flattens from 46 to 45 diopters. When the incisions are extended subsequently to a 3-mm optical zone, the patient develops 3 diopters of refractive change and the central cornea flattens from 46 to 42 diopters. It is apparent that as incisions approach the visual axis, corneal flattening increases. The wide separation of the central rings demonstrates that corneal flattening diminishes (long arrow) as the peripheral cornea is approached, producing a geometrical and optical corneal "knee" at the transition point between the central and peripheral cornea (short arrow).
150
OPHTHALMOLOGY •
FEBRUARY 1983 •
VOLUME 90 •
NUMBER 2
Table 2. Radial Keratotomy Eyebank Data-Effect of 16 Radial Incisions: Curvature Change in Diopters
3H Optical Zone
3 mm-limbus
Mean Range
low high
Eye #
12 15 11 15 9 10 17 3 8 16 18 24 28
3V Ring Number
5H
5V
9H
9V -1.37 -3.62
-7.50 -7.50 -8.75 -13.00 -11.25 -10.75 -11.75 -12.75 -12.12 -11.75 -14.00 -13.75
-8.00 -8.88 -8.25 -5.62 -9.62 -10.00 -9.50 -10.88 -11.62 -12.25 -12.75 -13.25
-2.88 -6.50 -6.75 -5.50 -5.75 -6.88 -7.25 -8.75 -7.50 -7.75 -9.50 -10.00 -11.00
-8.00 -6.50 -5.25 -5.00 -7.88 -7.50 -8.50 -7.50 -9.88 -9.50 -11.00 -9.50
-0.88 -2.62 -4.00 -3.75 +0.50 +0.50 -3.50 -4.25 -1.37 -2.25 -2.50 -4.75 -2.50
-2.50 -4.75 -3.50
-11.24 7.50 13.75
-10.05 5.62 13.25
-7.38 2.88 11.00
-8.00 5.00 11.00
-2.41 0.50 4.75
-2.90 1.37 4.75
When 16 radial incisions are placed in eyebank eyes, symmetrical central corneal flattening occurs of approximately 10.5 diopters. Less
flattening of the cornea is observed (Figs 8A, B). Flattening of the cornea is manifested by ring separation or broadening, indicating a longer (flatter) radius of curvature. Mattress sutures steepen the cornea within the suture area but also flatten the cornea directly under the penetration site of the suture into the corneal stroma.
5. THE LIMBAL CORNEA FLATTENS ADJACENT TO LOOSE SUTURES
-2.12 -3.88 -1.75 -1.75 -4.50 -2.12
midperipheral flattening is observed over the fifth and ninth rings respectively.
6. THE LIMBAL CORNEA STEEPENS ADJACENT TO TIGHT SUTURES (A) AND STEEPENS 180 0 AWAY, (B) AND FLATIENS 90 0 AWAY
Tight suturing of the limbus after cataract surgery may allow for control of preoperative corneal astigma-
(A) AND FLATIENS 180 0 AWAY, (B) AND STEEPENS 90 0 AWAY
The effect of corneal flattening may be advantageous or a disaster after cataract surgery, depending on the preoperative refractive error of the patient. Flattening of the limbal corneal incision with loose sutures (Caveat 1) is accompanied by elongation of the entire arc length of the cornea from the 12 to the 6 o'clock limbus, and elongation of the corneal chord or diameter (Fig 6). Flattening of the cornea in the meridian of loose sutures is routinely associated with concomitant steepening of the cornea 90° away, or in the horizontal meridian after cataract surgery. The normal 43 diopter spherical cornea may demonstrate wound dehiscence with a 38 diopter vertical meridian, (too flat) and a 48 diopter horizontal cornea (too steep). Relaxing incisions along the horizontal meridian would be inappropriate for superior wound gape. We have found it auspicious to reopen cataract incisions suffering from wound gape and resuturing the incisions with nonabsorbable sutures to compress the wound and add plus power to the involved cornea (Caveat 6) when necessary. 152
..
:
r\c/r='===:::~
r
\1
r
;
Fig 5. Caveat 3. Development of midperipheral knee after radial keratotomy is related to flattening of central cornea with less flattening of the peripheral cornea.
ROWSEY •
KERATOREFRACTIVE SURGERY
7. THE CORNEA FLATTENS OVER WEDGE RESECTION OR TUCK
AGAINST THE RULE
"....... I1>I'IU
Fig 6. Caveat 5. Diagram of wound gape with vertical arc (solid curve) and chord (dotted line) elongation following cataract surgery and horizontal shortening of the corneal radius of curvature, producing against-the-rule astigmatism.
tism. Although the cornea flattens directly over any suture, the posterior vector force of tight sutures is compensated by a plus power short radius of curvature effect anterior (central) to the sutures themselves (Fig 7). The balloon corneal model of Troutman5 demonstrates this caveat well. Tight sutures at the limbus produce a short radius of curvature adjacent to the opposite limbus and a long radius of curvature 90° away. Many authors6- 8 feel that slight with-the-rule astigmatism is desirable in the one- to four-week postoperative period following cataract surgery. An immediate wound compression force producing 1.5 to 2.5 diopters of with-the-rule astigmatism resolves following limbal incision suturing, producing an anastigmatic cornea. If sutures traverse the visual axis, no cornea remains centrally for subsequent steepening. Therefore, the total effect on corneal power is due to the corneal flattening itself and no plus power can be induced by tight or loose suturing.
r ..... \ r
~
............
Fig 7. Caveat 6. Following cataract surgery, tight sutures produce wound compression and vertical corneal astigmatism. Although the cornea flattens and lengthens the radius of curvature (r) directly under the suture (arrow), it steepens anterior to the suture producing withthe-rule astigmatism or a short radius of corneal curvature (r) centrally.
Wedge resections may produce marked dioptric power shifts in the cornea. The cornea flattens over a wedge resection more than tight suture contraction may allow. The wider or larger the wedge resection, 8 the larger the degree of flattening that occurs over the wedge resection itself. Wedge resections often involve one full quadrant of the cornea to produce symmetry of plus power anterior to the wedge. Both wedge resection and corneal tucks have diminished effect closer to the limbus and an increased effect as the visual axis is approached. The flattening effect of tissue removal may be more permanent with wedge resection than with corneal tucking procedures. If tissue has filled the interstices of a cataract incision wound gape after one to two years of healing, a wedge resection in the original incision with removal of the proliferative tissue will allow the addition of plus power to the cornea by flattening the limbal area. Wedge resections or tucks may be used in a corneal transplant donor-recipient interface if 10 to 12 diopters of correction are required (wedge), or 5 to 10 diopters of power are needed (tuck).
8. THE CORNEA STEEPENS ANTERIOR TO WEDGE RESECTIONS OR TUCKS The surgeon's goal is plus power addition in the meridian of a wedge resection. If the cornea is flatter than 43 diopters and significant astigmatism is present, the surgeon may anticipate adding plus power in the axis of the cornea where a tuck or wedge resection is accomplished. No tissue is removed during the tuck procedure, but a circumferential incision is placed near the limbus or in clear cornea. The incision is closed by mattress sutures (Fig. 9). Inelastic prolene sutures may cut through the corneal bed, and corneal power may therefore dissipate following a tuck. Large plus power dioptric corrections are possible with wedge resections and less regression may be observed due to permanent structural alterations.
9. TISSUE REMOVAL PRODUCES CORNEAL FLATTENING OVER THE SITE OF TISSUE REMOVAL, WHETHER TRAUMATIC OR SURGICALLY INDUCED Self-healing lacerations of the cornea are frequently not repaired and may be missed by both the patient and the physician. However, any penetrating corneal laceration is associated with flattening and decreased corneal power (Caveat 1). If stromal tissue is extracted by trauma or debrided at the time of surgical repair, the subsequent tight suture required to seal the laceration is routinely associated with marked flattening of the cornea. If tight 153
OPHTHALMOLOGY • FEBRUARY 1983 • VOLUME 90 • NUMBER 2
Fig 8. Caveat 4. A corneal laceration that has been meticulously sutured produces corneal flattening whether the sutures are tight or loose. A horizontal traumatic incision of the cornea has been closed carefully and demonstrates the flattening directly over the incision itself (arrows) by both widening of ring separations and broadening of the rings themselves.
Fig 9. Left, the vector of a compression of a suture is demonstrated as a short chord "V" configuration under the suture itself. A patient who has relaxing incisions and a single compression tuck suture at 4:30 o'clock demonstrates persistence of the symmetrizing effect of the suture " V" in this area (arrow). This vector of compression may be relieved by suture removal at any time in the postoperative period and the induced astigmatism will diminish. If wound compression is not observed, no diminution of corneal astigmatism can be anticipated by suture removal. Fig 10. RighI, Caveats 9 and 10. Corneascope photograph of patient who has undergone tectonic patch graft at 8:00 (arrow) for corneal astigmatism due to traumatic tissue loss. Tissue addition to the cornea steepens the area of tissue addition and flattens the visual axis, thereby reducing corneal astigmatism.
sutures are necessary in the visual axis, then decreased corneal power is realized after surgery. If, however, tight sutures are required in the midperipheral or peripheral cornea, marked flattening in the area of repair is associated with concomitant steepening of the cornea centrally in the visual axis (Caveat 6). Thus, repair of cor154
neal lacerations is similar refractively to an uncontrolled wedge resection of the cornea (Caveats 7 and 8). Suture removal early in the postoperative period may be associated with release of the flattening effect and diminished astigmatism while risking wound dehiscence and leak. In the area of tissue loss, early suture removal does
ROWSEY •
KERATOREFRACTIVE SURGERY
not replace the tissue loss and a flat area of the cornea may persist similar to postinflammatory thinning of the cornea.
10. FULL-THICKNESS CORNEAL TISSUE ADDITION PRODUCES CORNEAL STEEPENING OVER THE SITE OF THE TISSUE ADDITION AND FLATTENS THE ADJACENT CORNEA Corneal stromal incisions with tissue addition corrects the flattening effect observed with traumatic tissue removal. The flat cornea represents a radius of curvature that is excessively long, due to compression. The convex surface of additional corneal tissue reduces the preoperative corneal curvature and corrects both the tissue site and simultaneously flattens the visual axis (Fig 10), essentially reversing the effect of the traumatic wedge resection. Tissue addition to a traumatically flattened area of the cornea may thereby permit the patient to return to contact lens use if necessary.
SUMMARY We have observed ten basic caveats of topographic and keratorefractive change following surgical or traumatic incision in the cornea. Placido disc imagery or
corneoscope photography with subsequent careful review of the dioptric power shifts have allowed documentation of the refractive alterations in these procedures. Planned reconstructive procedures to reduce myopia, hyperopia, or astigmatism take into account these basic caveats. The appropriate removal of sutures following cataract or penetrating keratoplasty surgery is facilitated by an understanding of these ten caveats.
REFERENCES 1. Rowsey JJ, Reynolds AE, Brown R. Corneal Topography: corneascope. Arch Ophthalmol1981; 99:1093-100. 2. Doss JD, Hutson RL, Rowsey JJ, Brown DR. Method for calculation of corneal profile and power distribution. Arch Ophthalmol 1981; 99:1261-5. 3. Rowsey JJ, Balyeat HD. Prelirninary results and complications of radial keratotomy. Am J Ophthalmol 1982; 93:437-55. 4. Dahlstrom R. Personal cornrnunication. 5. Troutman RC. Microsurgery of the Anterior Segment of the Eye, Vol. II. The Cornea: Optics and Surgery. St. Louis: CV Mosby, 1977; 26873. 6. Jaffe NS. Cataract Surgery and its Cornplications. St Louis: CV Mosby, 1976; 83-98. 7. Jaffe NS, Clayrnan HM. The pathophysiology of corneal astigmatisrn after cataract extraction. Trans Am Acad Ophthalmol Otolaryngol 1975; 79:615-30. 8. Barner SS. Surgical induction of corneal astigmatism; An experirnental study. Albrecht von Graefes Arch Klin Exp OphthalmoI1977; 201 :213-20.
155