Corneal Topography and Fluctuating Visual Acuity after Radial Keratotomy PETER J. McDONNELL, MD, DRUANN J. McCLUSKY, MD, JENNY J. GARBUS, COT

Abstract: A high-resolution photo keratoscope using computer graphics to model corneal topography was used on patients who had undergone radial keratotomy. After radial keratotomy, central optical zones are created that can be characterized as round, oval or band-like, or dumbbell-shaped or split. The dumbbell form of optical zone was associated with larger amounts of refractive and keratometric astigmatism than the round or band-like zones. The authors correlated the shape of the optical zone with the presence or absence of diurnal variation (fluctuation) in visual acuity. Of the 26 eyes studied, 11 expe­rienced fluctuation and 15 did not. Of those 11 eyes with fluctuating visual acuity, 10 (91%) had dumbbell-shaped or split optical zones and 1 (9%) had a round optical zone. Of the 15 eyes without fluctuation, 12 (80%) had round optical zones and 3 (20%) had band-like zones. The presence of a split or dumbbell-shaped optical zone after radial keratotomy indicates that the patient is likely to experience diurnal fluctuation of visual acuity. Ophthalmology 96: 665-670, 1989

fluctuation of visual acuity and refractive error from morning to evening after radial keratotomy has been re­ported to occur in 1.9 to 60% of cases. I-to The diurnal change reported is usually a myopic shift from morning to evening of approximately 0.5 diopters (D). I I Some pa­tients, however, experience a decrease in myopia or an increase in hyperopia from morning to evening; O'Day et al l2 reported an eye with a spherical equivalent of -2.87 in the morning and -1.25 in the afternoon. The diurnal changes in refractive error after radial keratotomy are presumed to relate to the severing of stromal collagen lamellae by the radial incisions; the continuity of these collagen fibrils is not restored during the healing process. 13

Diurnal fluctuation in refractive error and visual acuity may persist for up to 6 years after surgery (unpublished data). The biomicroscopic appearance of these radial in­cisions over several years shows gradual changes that are

Originally received: August 10, 1988. Revision accepted: December 8, 1988.

From the Doheny Eye Institute and the University of Southern California School of Medicine, Los Angeles.

Reprint requests to Peter J. McDonnell, MD, Doheny Eye Institute, 1355 San Pablo St, Los Angeles, CA 90033.

interpreted as representing remodeling of the corneal stroma within the wounds. 14

To our knowledge, two hypotheses have been advanced to explain the diurnal variation in refractive state. In one experimental study, rabbit corneas after radial keratotomy flattened in response to elevations of intraocular pressure (lOP): as lOP was increased from 10 to 20 mmHg, the corneal refractive power decreased by 0.50 D.15 A similar study by the same investigators in non-human primates, however, indicated that elevations in lOP in the physio­logic range do not significantly affect refractive status. 16 In a clinical study performed on patients in the Prospective Evaluation of Radial Keratotomy (PERK) Study, Schan­zlin et al17 found no relationship between lOP and re­fraction. Wyzinski and O'Dell I I suggested that corneal thickness changes might be responsible for refractive changes. In support of their hypothesis, they described a patient who, 1 year postoperatively, was refracted before and after a 25-minute swim in fresh water and showed a I-D hyperopic shift in each eye. MacRae et al18 examined patients 1 to 2 weeks after radial keratotomy and observed an increase in myopia between morning and evening as well as a decrease in mean corneal thickness from morning to evening. This report describes the morning-to-evening changes in visual acuity, refractive error, central corneal thickness, and corneal topography after radial keratotomy.

665

OPHTHALMOLOGY • MAY 1989 • VOLUME 96 • NUMBER 5

PATIENTS AND METHODS

We examined 26 eyes (15 patients) at least 2 years after eight-incision radial keratotomies, which were performed with diamond knives. Surgery was performed by one of two surgeons. The blade depth was set equal to 100% of the thinnest of four paracentral pachymetric readings. There were no intra- or postoperative complications in any of the patients. All patients agreed to return for morning and evening examinations as part of their yearly examination after surgery. The interval between surgery and examinations ranged from 2 to 6 years (average, 4 years).

Patients were examined before 8:00 AM and again after 7:00 PM. Uncorrected and best-corrected visual acuities were measured using the same visual acuity charts. Central keratometric power was determined using a single Bausch & Lomb keratometer. Morning and evening examinations and manifest refractions, pachymetric readings, and ap­planation tonometry were performed by the same ob­server.

At each visit, high-resolution photo keratoscopy using computer graphics was used to measure corneal topog­raphy.19 Using calibrated steel balls, this device, which provides graphic presentation of computer-analyzed ker­atoscope photographs,20 has been shown to provide a measure of curvature as accurate and reproducible as that obtained with the Bausch & Lomb keratometer (Leandro, CA).21 Eight examinations of the same eye were performed at the time of each visit to determine the reproducibility of the measurements.

At the time of the patients' examinations, they were questioned regarding the presence or absence of subjective changes in visual acuity from morning to evening.

An ultrasonic pachymeter (Kremer corneometer, Ac­cutome, Inc, Frazer, PA) set to the speed of sound in the cornea (1640 m/second) was calibrated on a test block and used to measure central corneal thickness 5 seconds after instillation of a single drop of proparacaine hydro­chloride 0.5% ophthalmic solution. Readings in microm­eters were taken until three readings were identical.

Correlations were made between the morning-to-eve­ning change in uncorrected visual acuity and the morning­to-evening changes in spherical equivalent, average ker­atometry, lOP, and central corneal thickness. In addition, the morning-to-evening change in spherical equivalent was correlated with changes in keratometry, lOP, and corneal thickness. Finally, the change in keratometry was correlated with the changes in lOP and pachymetry. Cor­relation coefficients are reported for these analyses.

Optical zones were characterized as round, band-like, or split by an observer without knowledge of the morning­to-evening changes in visual acuity or refractive error.

Statistical analysis of the influence of optical zone shape on the presence or absence of subjective complaints of fluctuating vision was performed using the chi-square with one degree of freedom.

666

RESULTS

The findings for the 26 eyes included in this study are shown in Table I. Subjective complaints of fluctuation in II eyes were reported. From morning to evening, four eyes gained one line of Snellen acuity, six eyes had no change in Snellen acuity, six eyes had a decrease in Snellen acuity of one line, two eyes had a decrease of two lines, two eyes had a decrease of thee lines, three eyes had a decrease of four lines, and three eyes had a decrease of more than four lines in Snellen acuity. For all eyes, the mean change in acuity was a loss of 1.7 lines and the median change in acuity was a loss of one line. Of the II eyes with subjective change in visual acuity from morning to evening, the average change in acuity was a decrease of about 3.5 lines, ranging from an increase of one line to a decrease of eight lines.

The average change in spherical equivalent from morning to evening was -0.25 D (range, -1.62 to + 1.25 D). The absolute change in spherical equivalent from morning to evening averaged 0.73 D in those eyes with subjective complaints of fluctuating vision and 0.32 Din those eyes without subjective change.

Average keratometry values differed from morning to evening by less than or equal to 0.25 D in 18 (69%) of the eyes. In eight eyes, the average keratometry value in­creased between 0.26 and 0.50 D from morning to eve­ning. Most eyes were steeper vertically than horizontally (Le., with-the-rule astigmatism), but against-the-rule astigmatism was seen in three eyes.

The mean change in lOP from morning to evening was a decrease of 1.5 mmHg (range, -4 to + I mmHg).

Central pachymetry readings were unchanged from morning to evening in 15 (58%) of the 26 eyes. Eight eyes (31 %) had a decrease in central corneal thickness of 0.0 I mm, and three eyes had an increase in central corneal thickness of 0.01 mm.

Results of examination of the color-coded topographic map of each cornea showed an area of central flattening with a relatively steeper periphery (Figs 1-4). These areas of central flattening, or effective "optical zones," could be characterized as round (Fig I), band-like extending almost from limbus to limbus within the interpalpebral zone (Fig 2), and dumbbell-shaped or split (Fig 3). Com­parison of multiple studies of the same eye demonstrated a standard deviation of ±0.2 D, similar to that reported by others.21

As shown in Table 2, correlation coefficients did not indicate a statistically significant relationship between the morning-to-evening changes in uncorrected visual acuity and the changes in spherical equivalent, average kera­tometry, lOP, or central corneal thickness. Also, there were no statistically significant correlations between the change in spherical equivalent and the change in average keratometry, lOP, or pachymetry. Finally, the changes in central keratometry were not predicted by the changes in lOP or central corneal thickness.

McDONNELL et aI • TOPOGRAPHY AND FLUCTUA TlNG ACUITY

Table 1. Fluctuating Visual Acuity after Radial Keratotomy

Change in Change in Change in Central

Snellen Spherical Cylinder in Keratometric Measurement (0) Change in Corneal Shape Cornea Fluctu- Acuity Equivalent Spectacle lOP Thickness of Optic

No. ation (lines) (0) Plane (0) Morning Evening (mmHg) (mm) Zone

1 Yes - 8 -1.62 0.75 37.75@6/37.37 38.25@176/37.75 - 1 0 Split 2 Yes -5 -1.50 -1.00 38.87@85/38.25 39.12@82/38.62 0 -0.01 Split 3 Yes -2 -0.76 0.50 40.75@128/40.12 40.62@158/40.oo 0 0.01 Split 4 No -2 -0.63 0 40.50@60/40.25 40.50@48/40.25 -4 -0.01 Round 5 Yes +1 0.12 1.25 39.75@98/39.12 40.37@107/39.25 -3 -0.01 Split 6 No 0 -1 .24 0.50 39.00@loo/38.25 39.37@111/38.50 0 0 Round 7 No -1 -0.25 0.75 38.87@85/38.12 39.12@95/38.37 0.1 0 Round 8 Yes -4 -0.37 075 40.00@85/39.50 40.50@76/39.50 0 -0.01 Split 9 No - 1 0 1.50 41 .75@97/39.75 42.12@96/39.62 -1 0 Band

10 Yes -4 1.00 1.00 41 .87@116/40.75 41.37@96/40.62 -4 0 Split 11 Yes -6 1.25 0.50 37.37@100/37.00 38.75@94/38.12 0 0 Round 12 No -1 0.38 0.50 3925/39.50 39.50/39.50 -1 0 Round 13 Yes - 3 -0.50 1.75 41 .25@91/39.25 40.87@106/39.62 -1 0 Split 14 Yes -3 -0.25 1.50 41 .25@94/39.50 41 .37@98/39.62 -2 0.01 Split 15 No -1 -0.32 0.75 38.12@84/37.75 39.12@86/38.50 -1 0 Round 16 No 0 -0.50 0.75 40.37@104/39.50 41 .12@113/40.50 -2 -0.01 Round 17 Yes -4 -0.62 0.75 38.25@107/37.00 37.25@73/37.oo -2 0 Split 18 No -1 0.25 0.50 37.50@76/36.87 37.oo@76/36.62 -1 0 Round 19 No 0 0.25 0 37.75@85/37.62 38.37@111/37.87 -3 -0.01 Band 20 No +1 0 1.25 39.37@86/37.75 38.62@loo/37.87 -4 0 Band 21 No +1 -0.37 1.25 39.75@24/38.00 39.50@18/38.50 -1 0.01 Round 22 No 0 0.12 0.75 40.62@173/39.87 40.25@174/39.75 - 1 0 Round 23 No -1 -0.37 0.50 39.87@102/39.00 39.75@86/39.12 -2 -0.01 Round 24 Yes 0 -0.25 1.00 38.12@78/37.25 38.37@82/37.37 - 2 0 Split 25 No +1 -0.12 0.50 39.87@87/39.25 41.75@25/39.87 0 0 Round 26 No 0 0 0.25 39.12@84/38.87 39.87@8/39.37 -1 -0.01 Round

o = diopters; lOP = intraocular pressure.

The shape of the optical zone was correlated with the presence or absence of complaint of diurnal fluctuation in uncorrected visual acuity (Table 3). Of those II eyes with subjective fluctuating visual acuity, 10 (91%) had dumbbell-shaped or split optical zones and one (9%) had a round optical zone (Fig 4). Of the 15 eyes without sub­jective fluctuation, none had split optical zones; all 15 had round (12 eyes) or band-like (3 eyes) optical zones. The presence of a dumbbell-shaped or split optical zone was strongly correlated with the presence of the complaint of fluctuating visual acuity (P < 0.(01).

but the central cornea was relatively unaffected, and fluc­tuating vision was not reported. In a single eye with a round, but inferiorly decentered optical zone, however, the peripheral steepening was accompanied by fluctuating vision (Fig 4).

Round or band-like optical zones were associated with less cylinder in the spectacle plane (average, 0.62 D; range, 0-1.50 D) than were the dumbbell-shaped zones (average, 0.91 D; range, 0.05-1.75 D). Similarly, keratometric astigmatism tended to be greater with dumbbell-shaped zones (average, 0.91 D; range, 0.37-2.00 D) than with round or bell optical zones (average, 0.82 D; range, 0-1.87 D).

In all eyes examined, including an unoperated eye (Fig 5), topographic changes could be demonstrated from morning to evening. In patients with dumbbell-shaped or split optical zones, the central flattened isthmus eventually steepened, with a corresponding decrease in uncorrected visual acuity in eyes undercorrected after surgery (Fig 3). These changes were seen up to 6 years after surgery. In eyes with round (Fig I) or band-like (Fig 2) optical zones, peripheral steepening was seen from morning to evening

In unoperated eyes the cornea flattens from center to periphery. One normal eye, between 7 AM and 7 PM, showed overall flattening of the cornea in contrast to the steepening typically seen after radial keratotomy (Fig 5); this eye did not exhibit any fluctuation in uncorrected visual acuity. In some normal eyes with high astigmatism (>3 D), we have observed a dumbbell-shaped area of steepening above and below the corneal apex, but these have not shown diurnal fluctuation of topography nor have they been associated with diurnal changes in acuity or refractive error. None of our post-radial keratotomy patients, however, had preoperative astigmatism in excess of 1.25 D.

DISCUSSION

The lack of correlation between changes in uncorrected visual acuity, spherical equivalent, and keratometry seen in our study are consistent with findings of other inves­tigators who have examined the possible relationships be­tween these variables. 16 In a large series of the PERK Study

667

OPHTHALMOLOGY • MAY 1989 • VOLUME 96 • NUMBER 5

patients, Schanzlin et al'7 found that changes in keratom­etry did not correlate with changes in refractive error. This discrepancy does not mean that diurnal refractive changes are not due to changes in corneal topography; rather the keratometer, because it measures only an av­erage curvature of the central 3 mm of the cornea, may not provide a sufficiently detailed picture of the diurnal changes occurring in the corneal topography. Our study, using computer-assisted corneal topographic analysis,

668

Fig I. Top left. color-coded topographic map of cornea taken in the morning 3 years after radial keratotomy. Central cornea has a flat, round, optical zone with steepening toward the pe­riphery. Top right. same cor­nea as top left. examined in evening. Central optical zone shows minimal change; pe­ripheral cornea shows mild steepening. Central corneal power is 0.3 0 less than the morning examination. Fig 2. Large, flat. optical zone fills most of interpalpebral space. Minimal change occurs be­tween morning (second row left) and evening (second row right) examinations. Eye did not experience diurnal fluc­tuation in visual acuity. Fig 3. Morning (third row left), noon (third row right). and evening (bottom) examina­tions of eye with persistent diurnal fluctuation in visual acuity 6 years after radial keratotomy. Central dumb­bell-shaped. flat optical zone becomes split into two zones from morning to evening, and central corneal power increased 1.5 0 from morn­ing to evening.

suggests that specific characteristics of the central flattened optical zone are predictive of whether or not a patient will experience diurnal fluctuations in visual acuity. Pa­tients with large, round, central optical zones proved to be largely immune to the problem of diurnal fluctuation. Patients whose zones of flattening were dumbbell in shape or split, such that two areas of flattening were encroached upon or separated by a steeper area were likely to complain of diurnal fluctuation in visual acuity. A prospective study

McDONNELL et al • TOPOGRAPHY AND FLUCTUATING ACUITY

Fig 4. Morning (top left) and evening (top right) examina­tions of cornea with diurnal fluctuation 3 years after ra­dial keratotomy. Round, flat optical zone is decentered in­feriorly; central corneal power increases from 39.8 to 40.3 0 from morning to eve­ning. Fig 5. Morning (bottom left) and evening (bottom right) examinations of nor­mal, unoperated cornea of moderately myopic individ­ual. Corneal flattening is seen from morning to evening; central corneal power de­creases from 44.7 to 44.1 D. Patient detected no fluctua­tion in acuity with spectacles in this eye.

of corneal topography before and after radial keratotomy is under way to determine if there are any preoperative topographic features predictive of postoperative diurnal fluctuation.

The explanation for the variation in the appearance of the areas of flattening is not obvious. All patients were operated on according to the same technique, with cre­ation of an eight-incision radial keratotomy using a dia­mond blade set to 100% of the thinnest of the paracentral corneal thicknesses measured intraoperatively. In none of the patients were there any recognized intra- or post­operative complications, such as microperforations or in­correct marking of the visual axis. Postoperatively, find­ings of slit-lamp examination did not show any obvious asymmetry in incision depth.

The etiology of diurnal fluctuations in visual acuity after radial keratotomy remains unknown. Despite the experimental evidence in rabbits that elevations in lOP can alter anterior corneal curvature,15 there was no trend in our study to suggest that changes in lOP were respon­sible for the morning-to-evening changes noted in the corneal topography of these patients. In primates after radial keratotomy, lOP does not affect corneal curvaturel6

;

similarly, Schanzlin et al17 found no relationship between lOP and refractive error in humans.

We obtained measurement of central corneal thickness at the time of morning and evening examinations, and again observed no trend to suggest that the alterations in corneal curvature were on the basis of changes in corneal thickness. These data are in disagreement with the findings of MacRae et al,18 who found large diurnal changes in central corneal thickness. However, there are two major differences between our study and that of MacRae et al.

Table 2. Correlation Coefficients of Morning-to-evening Changes

Dependent Variables

Change in Change in Change in Average

Independent Snellen Spherical Keratometric Variables Acuity Equivalent Power

Change in spherical equivalent 0.198

Change in average keratometric power 0.280 0.318

Change in lOP 0.146 0.151 0.207 Change in central

corneal thickness 0.034 0.040 0.053

Table 3. Fluctuating Visual Acuity and Optical Zone Shape*

Fluctuation

Present Absent

Split Optical Zone (no. of eyes)

10 o

* Chi-square = 18.5; P < 0.001.

Round or Band Optical Zone

(no. of eyes)

1 15

All the patients in our study were examined at least 12 months postoperatively, whereas the study by MacRae et al involved patients who were only 1 to 2 weeks postop­erative. Although our patients were seen at approximately

669

OPHTHALMOLOGY • MAY 1989 • VOLUME 96 • NUMBER 5

the same time in the morning as were MacRae et al's, we did not ask our patients to tape their lids shut, so it is possible that some stromal dehydration occurred before we saw our patients for their morning examinations. De­spite this, we clearly documented changes in corneal cur­vature, refractive error, and visual acuity that were not accompanied by changes in central corneal thickness. An alternative explanation for the changes we observed might be alteration in lid position (closed during sleep, at upper limbus during waking hours) which might induce topo­graphic changes in a cornea somewhat destabilized as a result of eight radial incisions. Further clinical and basic research will be necessary to explain the diurnal topo­graphic changes that have been documented after radial keratotomy.

REFERENCES

1. Neumann AC, Osher RH, Fenzl RE. Radial keratotomy: a compre­hensive evaluation. Doc Ophthalmol1984: 56:275-301.

2. Bores LD, Myers W, Cowden J. Radial keratotomy: an analysis of the American experience. Ann Ophthalmol 1981; 13:941-8.

3. Cowden JW, Bores LD. A clinical investigation of the surgical correction of myopia by the method of Fyodorov. Ophthalmology 1981 ; 88: 737-41.

4. Gelender H, Flynn HW Jr, Mandelbaum SH. Bacterial endophthalmitis resulting from radial keratotomy. Am J Ophthalmol1982; 93:323-6.

5. Hoffer KJ, Darin JJ, Petitt TH, et al. UCLA clinical trial of radial kera­totomy: preliminary report. Ophthalmology 1981; 88:729-36.

6. Deitz MR, Sanders DR, Marks RG. Radial keratotomy: an overview of the Kansas City study. Ophthalmology 1984; 91 :467-78.

7. Rowsey JJ, Balyeat HD. Preliminary results and complications of radial keratotomy. Am J Ophthalmol1982; 93:437-55.

670

8. Rowsey JJ, Balyeat HD. Radial keratotomy: preliminary report of com­plications. Ophthalmic Surg 1982; 13:27-35.

9. Kremer FB, Marks RG. Radial keratotomy: prospective evaluation of safety and efficacy. Ophthalmic Surg 1983; 14:925-30.

10. Sarno EM, Smith RE, Schanzlin OJ. Comparison of clinical results following radial keratotomy, extended-wear contact lenses, and myopic keratomileusis. Int Ophthalmol Clin 1983; 23(3):167-92.

11. Wyzinski P, O'Dell LW. Diurnal cycle of refraction after radial keratot­omy. Ophthalmology 1987; 94:120-4.

12. O'Day OM, Feman SS, Elliott JH. Visual impairment following radial keratotomy: a cluster of cases. Ophthalmology 1986; 93:319-26.

13. Ingraham HJ, Guber 0, Green WR. Radial keratotomy: clinicopathologic case report. Arch Ophthalmol1985; 103:683-8.

14. Steinberg EB, Waring GO, Wilson LA. Slitlamp microscopic study of corneal wound healing after radial keratotomy in the PERK Study. ARVO Abstracts. Invest Ophthalmol Vis Sci 1985; 26(Suppl):203.

15. Busin M, Yau CoW, Avni I, et al. The effect of changes in intraocular pressure on corneal curvature after radial keratotomy in the rabbit eye. Ophthalmology 1986; 93:331-4.

16. Busin M, Arffa RC, McDonald MB, Kaufman HE. Change in corneal curvature with elevation of intraocular pressure after radial keratotomy in the primate eye. CLAO J 1988; 14:110-2.

17. Schanzlin OJ, Santos VR, Waring GO III, et al. Diurnal change in re­fraction, corneal curvature, visual acuity, and intraocular pressure after radial keratotomy in the PERK Study. Ophthalmology 1986; 93:167-75.

18. MacRae SM, Rich LF, Bedrossian RH. Diurnal variation of vision fol­lowing radial keratotomy. ARVO Abstracts. Invest Ophthalmol Vis Sci 1986; 27(Suppl):1S.

19. Gormley OJ, Gersten M, Koplin RS, Lubkin B. Corneal modeling. Cor­nea 1988; 7:30-S.

20. Maguire LJ, Singer DE, Klyce SO. Graphic presentation of computer­analyzed keratoscope photographs. Arch Ophthalmol1987; 10S:223-30.

21. Hannush SB, Waring GO III. Computer assisted corneal topography: accuracy and reproducibility with three instruments. ARVO Abstracts. Invest Ophthalmol Vis Sci 1988; 29(Suppl):389.