Embed Size (px)
Citation preview
Morning ... to ... evening Change in Refraction, Corneal Curvature, and Visual Acuity 11 Years after Radial Keratotomy in the Prospective Evaluation of Radial Keratotomy Study
Peter]. McDonnell, MD/ Azhar Nizam, MS,2 Michael]. Lynn, MS,2 George O. Waring III, MD, 2 the PERK Study Group*
Purpose: Previous reports demonstrate morning-to-evening changes in ophthalmic measurements at 3 months, 1 year, and 4 years after radial keratotomy. The authors determine whether diurnal change in refractive error persists 11 years after radial keratotomy surgery in the Prospective Evaluation of Radial Keratotomy (PERK) study.
Methods: Seventy-one patients were examined in the morning and evening a mean of 11.1 ± 0.6 years (range, 10-12.7 years) after undergoing radial keratotomy under a standardized protocol using a diamond blade.
Results: Between the morning and evening examinations, the mean change in the spherical equivalent of refraction was a 0.31 ± 0.58-diopter (0) increase in minus power in first eyes. Thirty-six (51 %) eyes had an increase in minus power of the manifest refraction of 0.50 to 1.62 0; 22 (31 %) had a change in refractive cylinder power of 0.50 to 1.25 0; 9 (13%) had a decrease in uncorrected visual acuity of two to seven Snellen lines; and 25 (35%) showed central corneal steepening measured by keratometry of 0.50 to 1.94 O. Two (3%) eyes lost two lines of spectacle-corrected visual acuity, whereas one (1 %) eye gained two lines. In patients whose both eyes underwent surgery, a high degree of symmetry was observed in morning-to-evening refractive change.
Conclusion: In some patients after radial keratotomy, morning-to-evening change of refraction and visual acuity persists for at least 11 years, although in most patients the magnitude of this change is small. Thus, diurnal fluctuation may be a permanent sequela of radial keratotomy in some individuals. Ophthalmology 1996;103:233-239
Originally received: October 14, 1994. Revision accepted: January 30, 1995. I Doheny Eye Institute, University of Southern California School of Medicine, Los Angeles. 2 Emory University, Atlanta. Presented at the American Academy of Ophthalmology Annual Meeting, San Francisco, Oct/Nov 1994.
In previous studies, we documented changes from morning to evening in manifest refractive error, uncorrected visual acuity, and average central keratometry at both 3 months and 1 year! and between 2.5 and 4 years2 after radial keratotomy in the Prospective Evaluation of Radial Keratotomy (PERK) Study. A subset of the patients in
The authors have no proprietary interest in the information presented in this article. Supported by National Eye Institute/National Institutes ofHeaIth grants EY03752 and EY03761, Bethesda, Maryland.
Dr. McDonnell is a William and Mary Greve International Research Scholar, Research to Prevent Blindness, Inc, New York, New Yark.
Reprint requests to Peter J. McDonnell, MD, 1450 San Pablo St, Los Angeles, CA 90033.
233
Ophthalmology Volume 103, Number 2, February 1996
the PERK Study also showed an increasing effect of the surgery (i.e., further change in the refractive error in the hyperopic direction) over years after surgery.2,3 We previously examined 63 patients at 3 months and 46 patients at 1 year in the morning and evening. Of the surgical eyes (1 per patient) examined at 1 year, 42% had an increase in minus power ofthe manifest refraction of 0.50 to 1.25 diopters (D) from morning to evening; 21 % had a decrease in uncorrected visual acuity of two to four Snellen lines; and 35% showed steepening of the central cornea by 0.50 to 1.25 D. The proportion of eyes that changed at 3 months and 1 year was similar, indicating that many corneas had not stabilized by 1 year after radial keratotomy. These patients were re-examined between 2.5 and 4 years after surgery; 31 % of the eyes had an increase in minus power of the spherical equivalent of the manifest refraction of 0.50 to 1.50 D and 19% had a decrease in uncorrected visual acuity of two to five Snellen lines. The magnitude of morning-to-evening change did not follow a consistent pattern; some eyes had greater fluctuation at 2 to 4 years than at 1 year. These data raise the question of whether diurnal fluctuation may be a permanent complication of radial keratotomy in some eyes or whether the fluctuation decreases as wound healing of the radial corneal incisions is completed many years after the surgery. To address this concern, morning and evening examinations were performed at 11 years after surgery in the PERK Study.
Patients and Methods
Study Population The PERK Study is a nine-center clinical trial of a standardized technique of radial keratotomy funded by the National Eye Institute. The design of the studl and major results at 5 and 10 years after surgery3,5 have been published previously. Participants in the current study were enrolled from all nine PERK clinical centers.
Two months after radial keratotomy surgery, the clinical coordinators at five of the nine centers telephoned their patients and asked if they had any changes in vision over the course of the day. Those who answered "yes" were encouraged to participate in the study of their symptoms. Thus, these participants were not selected randomly, but were self-selected to document changes in ophthalmic measurements in patients reporting fluctuating vision. Patients who had participated in the study at 3 months (n = 63) or at 1 year (n = 46) and at 2 to 4 years (n = 52) were asked to participate in a 10-year study to determine whether measurable changes persisted. In addition, patients were recruited at the four centers that had not participated in previous morning-to-evening studies. These patients were self-selected from the entire population of patients at these four centers. Forty of the 71 patients at 10 years had been examined as part of the previous morning-to-evening studies.
Surgical Technique The standardized protocol for the PERK surgical technique has been described previously.4 Briefly, the pre-
234
operative spherical equivalent of the cycloplegic refraction determined the diameter of the central clear zone: -2.00 to -3.12 D, 4.0 mm; -3.25 to -4.37 D, 3.5 mm; and -4.50 to -8.00 D, 3.0 mm. The diamond blade of a micrometer knife was extended to a length equal to 100% of the thinnest of four paracentral intraoperative ultrasonic pachometry measurements, and the length was verified on a coin gauge block. The surgeon made eight freehand single-pass, centrifugal radial incisions spaced equidistantly around the cornea. Of the 71 surgical eyes included in this study, 17% received a 4.0-mm clear zone, 31 % received a 3.5-mm clear zone, and 52% received a 3.0-mm clear zone.
Examination Procedure
Each patient was examined between 5:00 and 11 :00 AM and again between 4:00 and 8:30 PM of the same day, between 10 and 13 years after surgery. The examination hours were extended beyond those in previous morningto-evening studies to encourage patient participation. All morning-to-evening measurements were separated by at least 5.5 hours. One certified PERK coordinator from each center performed morning and evening examinations in the same standardized examination room using the same equipment. The following measurements were obtained: the manifest refraction using the fogging technique, the uncorrected and spectacle-corrected visual acuity using standardized PERK visual acuity charts, the central keratometric power using a calibrated keratometer, the central corneal thickness with an ultrasonic pachometer, and the intraocular pressure with a Goldmann applanation tonometer. These examination techniques have been described previously.4 The change from morning to evening in both the spherical equivalent refractive error and the cylinder power is reported.
The visual acuity was measured using modified BaileyLovie charts with five Sloan letters of approximately equal difficulty on each of 14 lines that ranged from 20/200 to 20/10. The visual acuity was recorded as the total number ofletters read. To quantify the Snellen acuity, we determined the position on the chart where the patient would have stopped if the total number ofletters had been read in sequence with none missed, and the final Snellen acuity was the line on which at least three letters would have been read.
Data Management and Analysis
Data for this study were recorded on standardized forms and managed in accordance with previously established PERK Study procedures.
The II-year morning-to-evening change was calculated as the evening measurement minus the morning measurement for all ophthalmic variables on the first surgical eye of each patient. For each variable, the distribution of the change was described by calculating the mean and standard deviation as well as the percentages of values within ranges of interest for that variable. The fluctuation in cylinder power of the manifest refraction was calculated
McDonnell et al . Diurnal Changes at 11 Years in PERK
as the absolute value of the evening measurement minus the absolute value of the morning measurement.
The II-year results were compared with the results at 2 to 4 years (mean, 3.5 years).2 The percentages of eyes that were stable (i.e., those that had no meaningful change) at both time points, stable at one time point, and unstable at both time points were calculated. For the manifest refraction and average central keratometric power, stability was defined as a change ofless than 0.5 D from morningto-evening. For uncorrected visual acuity, stability was defined as a change of one line or less from morning-toevening. Paired Student's t tests were performed to compare the mean morning-to-evening changes at the two time points for each of these variables.
For patients who had surgery on both eyes, the symmetry of fluctuation in manifest refraction and average central keratometric power was examined. Paired Student's t tests were performed to compare the mean fluctuation in the first- and second-surgical eyes. The percentages of patients for whom the difference in morningto-evening change between the two eyes was 0.5 D or less and 1.0 D or less were calculated. Data for patients who had a high magnitude of change in their first-surgical eye were examined to determine if they also had a high magnitude of change in their second-surgical eye. The symmetry of fluctuation also was examined for the nine patients who had unilateral surgery.
Correlation coefficients were calculated to examine the relation between the change from morning to evening in uncorrected visual acuity, manifest refractive error, central keratometric power, and intraocular pressure. Correlation coefficients also were computed between the change in refractive error and the refractive error before surgery, the refractive error after surgery, the patient's age, the preoperative intraocular pressure, and the depth of the incision scar to examine factors that might explain or predict fluctuation in vision.
The median change in cycloplegic refraction from 6 months to 10 years was calculated for the 71 eyes included in this study. The correlation between the 6-month to 10-year change in cycloplegic refraction and the II-year morning-to-evening fluctuation in manifest refraction was calculated. A chi-square test was performed to investigate the association between progressive hyperopia (defined as a decrease in the minus power of the cycloplegic refraction of at least 1.0 D from 6 months to 10 years) and morningto-evening fluctuation (defined as a change in manifest refraction of at least 0.5 D at 11 years).
Results
Population and Time of Examination
Seventy-one first-surgical eyes included in this study were measured both in the morning and in the evening at least 10 years after radial keratotomy surgery. Measurements were taken after 10 years but before 12.7 years (mean postoperative interval, 11.1 ± 0.6 years). The length of follow-up time after surgery was not related to the morn-
ing-to-evening change in uncorrected visual acuity (r = -0.07), manifest refractive error (r = -0.14), or average central keratometric power (r = 0.16). Thus, we pooled data from all patients without considering the time after surgery.
The spherical equivalent of the manifest refraction measured at the morning examination varied from -4.00 to +5.50 D. Of these eyes, 16 (23%) were myopic (-0.62 to -4.00 D), 25 (35%) were hyperopic (+0.62 to + 1.25 D), and 30 (42%) were emmetropic (-0.50 to +0.50 D).
Morning,to,evening Change
Refractive Error. The median change in refractive error was an increase in the minus power of 0.50 D from morning to evening (range, 1.62-D increase in minus power to 3.00-D decrease in minus power). The minus power increased an average of 0.31 D (standard deviation, 0.58 D) from morning to evening. Thirty-eight (54%) of the 71 eyes had a change in refractive error of 0.50 D or more, and for all except 2 eyes (3%) the change was an increase in minus power at the evening examination (Table 1).
Cylinder Power. The change in cylinder power of the manifest refraction from morning to evening was 0.25 D or less for 49 (69%) eyes. Fifteen (21 %) eyes were more astigmatic in the evening by 0.50 to 1.0 D. Seven (10%) eyes were more astigmatic in the morning by 0.5 to 1.25 D. The axis of the cylinder did not change (mean change, 0.71°; P = 0.92).
The change in the spherical equivalent was more often due to a change in the spherical rather than the cylindrical component. Of the 38 eyes whose spherical equivalent changed by 0.50 D or more, 33 (87%) had a change in the spherical power of 0.50 D or more (median change in spherical power was an increase in minus power of 0.75 D, ranging from a 1.75-D increase in minus power to a 2.50-D decrease in minus power). Of the 38 eyes whose spherical equivalent changed by 0.50 D or more, 23 had a change of 0.25 D or less in the cylinder power.
Uncorrected Visual Acuity. The visual acuity without refractive correction decreased an average of2.03 letters (standard deviation, 8.17 letters) from morning to evening (range, 35 letters lost at the evening examination to 20 gained). There was a meaningful change of two to
Table 1. Morning,to,evening Change in Manifest Refraction at 11 Years (71 eyes)
Change in Minus Power of Manifest Refraction
Increase 1.12-1.62 D Increase 0.50-1.00 D <0.50 D Decrease 1.00 D Decrease 3.00 D
D = diopter.
Percent of Eyes
6 45 46
1 1
235
Ophthalmology Volume 103, Number 2, February 1996
Table 2. Morning-to-evening Change in Uncorrected Visual Acuity at 11 Years (71 eyes)
Change in Snellen Visual Acuity
Loss 7 lines Loss 2-4 lines Loss 1 line No change Gain 1 line Gain 2-4 lines
Percent of Eyes
11 27 39 11 11
seven lines for 16 (23%) eyes; for most eyes, the change was a loss in uncorrected visual acuity (Table 2).
Spectacle-corrected Visual Acuity. The spectaclecorrected visual acuity decreased an average of 0.31 letters (standard deviation, 3.2 letters) from morning to evening. Two (3%) eyes lost two lines and one (1 %) gained two lines of visual acuity.
Keratometric Power. The central cornea was steeper at the evening examination by an average of 0.39 D (standard deviation, 0.41 D). Twenty-five (35%) eyes had corneas that were steeper at the evening examination by 0.50 to 1.94 D (Table 3).
Intraocular Pressure. There was an average decrease of 1.3 mmHg (standard deviation, 2.6 mmHg) in the intraocular pressure from morning to evening.
Symmetry of Morning-to-evening Change in Patients with Bilateral Surgery
In the 61 patients with surgery performed in both eyes, the magnitude and direction of morning-to-evening changes in refraction and keratometric power were compared. For both measures, a high degree of symmetry was apparent (Fig 1). The difference in change in the manifest refraction between the two eyes averaged 0.03 ± 0.56 D (no significant difference; P = 0.71), and was less than or
equal to 0.5 Din 50 (82%) patients and less than or equal to 1.0 D in 58 (95%) patients. For keratometric power, the difference in morning-to-evening change between the patients' two eyes averaged 0.13 ± 0.46 D (statistically significant, P = 0.03), with 51 (84%) and 58 (95%) patients having the two eyes within 0.5 and 1.0 D, respectively. Seventeen patients had more than 0.5 D of change in their first-surgical eye at 11 years. Of these 17 patients, 14 (82%) had a change of more than 0.5 D in their secondsurgical eye. Thus, a high magnitude of diurnal fluctuation in one eye is predictive of a high magnitude of fluctuation in the second eye.
Symmetry of Morning-to-evening Change in Patients with Unilateral Surgery
Only nine patients had surgery performed in only one eye, seriously limiting the ability to look for relations between diurnal changes between surgical and nonsurgical eyes. The mean change in manifest refraction from morning to evening in the nine surgical eyes was -0.29 D (standard deviation, 0.21 D). Eight of these eyes showed an increase in minus power from morning to evening. The mean change in the nine nonsurgical eyes was -0.06 D (standard deviation, 0.47 D), with only five of nine eyes showing an increase in minus power. Similarly, surgical eyes showed a greater increase in average keratometric power (mean, 0.48 ± 0.61 D) than nonsurgical eyes (mean, 0.15 ± 0.28 D).
Morning-to-evening Change at 3.5 Years Compared with Morning-to-evening Change 11 Years after Radial Keratotomy
Of the 71 eyes examined in the current study, 31 also were examined both morning and evening at 3.5 years after surgery. The changes from morning to evening in the current measurements were compared with the changes at 3.5 years for these 31 eyes.
Change in Manifest Refractive Error. Ten (32%) of the 31 eyes had less than 0.5 D of change in refractive
Table 3. Morning-to-evening Change in Cylinder Power, Central Keratometric Power, and Intraocular Pressure at 11 Years (71 first eyes)
Variable
Cylinder power (D) (n = 71; change: mean ± SD,
0.07± 0.39-D increase) Central keratometric power (D) (n = 71; change: mean ± SD,
0.39 ± 0.41 D steeper) Intraocular pressure (mmHg) (n = 70; change: mean ± SD,
1.3 ± 2.59-mmHg increase)
D = diopter; SD = standard deviation.
236
Decrease
3.00-7.33 30%
Decrease
0.5-1.25 10%
1.00-2.67 26%
Minimal Change
<0.5 69%
<0.5 65%
<1.00 29%
Increase
0.5-1.00 21%
0.5-1.9 35%
1.00-2.00 11%
Increase
5.00-7.67 4%
McDonnell et al . Diurnal Changes at 11 Years in PERK
co c:
~ 2
1.5
0.5
o
-0.5
-1
-1.5
-2
/ /
/ /
/ /
/ /
/ /
/
/
/ /
/ /
/
/ /
· / / ~
/ .. /
/ /
/
/ /
/ / ,../ ... / /
/ . . /
/ .. / · / • • / . ~
/ / .
/ . •
/
/
/ . /
/
/ /
/ /
/ /.
/ /
/ /
/ /
/ /
/
n = 61 Difference in fluctuation between two eyes: <= 0.5 0 82% (within dashed lines) <= 1.00 95% <= 1.50 97%
Avg. difference: 0.03 0 (50=0.560)
Figure 1. Comparison of morning-to-evening changes in manifest refraction in the two eyes or bilaterally surgical patients. The changes were within 0.5 diopter for 82% of patients (points within dashed lines), within 1.00 diopter for 95%, and within 1.50 diopter for 97%.
-2.5 --+---,----r----r----,---,--,----,----,----,-------r
-2 -1.5 -1 -0.5 o 0.5 1.5 2 2.5 3 Morning to Evening Change in Manifest Refraction (D)
In First Operated Eye
error at both 3.5 and II years and therefore were considered stable at both time points. At both 3.5 and II years, seven (23%) eyes had a persistent increase in minus power of 0.5 to 1.5 D, whereas two (6%) had a persistent decrease in the minus power from 0.5 to 0.875 D. Six (19%) eyes changed by 0.5 to 1.5 D at 3.5 years but were stable at 11 years. Five (16%) eyes were stable at 3.5 years but showed fluctuation of 0.5 to 0.625 D at 11 years. The average changes at 3.5 and 11 years were similar (P = 0.31).
Change in Uncorrected Visual Acuity. There was no meaningful change (morning-to-evening change :0:; I line) for 19 (61 %) eyes at both time points. Persistent increases in uncorrected visual acuity of two to five lines occurred in two (6%) eyes. Seven (23%) eyes changed by two to four lines at 3.5 years but were stable at II years, whereas three (10%) eyes were stable at 3.5 years but changed by two lines at II years. The average changes at 3.5 and II years were similar (P = 0.10).
Change in Central Keratometric Power. There was no meaningful change (morning-to-evening change < 0.5 D) for 12 (39%) of the eyes at both time points. There was steepening of 0.5 to 0.935 D for four (13%) eyes at both times. Seven (23%) eyes were stable at 3.5 years but steepened by 0.56 to 1.94 D at II years; eight (26%) eyes steepened at 3.5 years by 0.5 to 1.375 D but were stable at II years. The average changes were statistically similar at the two times (P = 0.99).
Correlations among Ophthalmic Measurements The change from morning to evening in uncorrected visual acuity was unrelated to the change in refractive error
(r = 0.18) or the change in central keratometric power (r = -0.01). An increase in keratometric power was not associated with an increase in minus power of the manifest refraction (r = -0.10). There was no relation between the morning-to-evening change in intraocular pressure and the change in uncorrected visual acuity (r = 0.05), the change in manifest refraction (r = 0.08), or the change in central keratometric power (r = 0.08).
Factors Influencing the Magnitude of Morningto-evening Change
We did not identify any factors that indicated which eyes were likely to have the greatest change in refractive error from morning to evening. The amount of change in manifest refractive error did not correlate well with the refractive error before surgery (r = -0.02), the refractive error after surgery (r = -0.06), the patient's age (r = -0.18), the preoperative intraocular pressure (r = -0.14), or the depth of the incision scar (r = -0.07).
Relation between Progressive Hyperopia and Diurnal Fluctuation
The median change in cycloplegic refraction from 6 months to II years for the 71 patients in this study was 1.0 D in the direction of continued effect of the surgery (mean, 1.28 ± 1.38 D). Diurnal fluctuation (morning-toevening change of at least 0.5 D) was present in 20 (53%) of the eyes that had at least 1.0 D of hyperopic progression from 6 months to 10 years, and for 18 (55%) of the eyes
237
Ophthalmology Volume 103, Number 2, February 1996
that did not have such progression. This difference was not statistically significant (P = 00.872). The correlation between the 6-month to 10-year change in cycloplegic refraction and the morning-to-evening change in manifest refraction was weak but statistically significant (r = 0.38; P = 0.00 I), with greater morning-to-evening change being associated with greater long-term refractive progression.
Discussion
Two types of long-term changes in refraction have been reported after radial keratotomy: diurnal fluctuation and hyperopic shift. The PERK Study previously reported morning-to-evening (diurnal) fluctuation in uncorrected visual acuity, manifest refractive error, and central keratometric power between 3 months and 3.5 years after radial keratotomy surgery.I,2 The findings reported here demonstrate that morning-to-evening changes persisted 11 years after radial keratotomy in some eyes, with some having fluctuations ranging from 1.5 to 3.0 D. In most patients, the changes were modest, with 55 (77%) eyes having a difference between morning and evening uncorrected visual acuities of one line or less.
This study measured the amount of morning-to-evening variation among a group of patients who had fluctuating vision when questioned 2 months after surgery, as well as additional subjects who had not participated in earlier morning-to-evening studies who were available for this examination. Patients were not selected randomly, but were self-selected based on their visual symptoms and their willingness to participate in the study. We included self-selected patients because our purpose was to document morning-to-evening changes and to follow these patients to determine if the changes persisted at 11 years. It was not our intent to evaluate the prevalence or magnitude of morning-to-evening fluctuation in the total PERK population. The refractive and visual acuity changes that these patients had should not be construed as representative of the entire PERK population.
One objective for the current study was to determine whether the refractive state had stabilized, particularly in those patients examined 3.5 and 11 years after surgery. Of the 71 eyes examined both morning and evening at 11 years, 51 % had an increase in the minus power of the manifest refraction of 0.50 to 1.62 D, 13% had a decrease in uncorrected visual acuity by two to seven Snellen lines, and 35% showed central corneal steepening by 0.50 to 1.94 D. These data indicate that approximately one third to one half of these corneas had not stabilized, although at least 10 years had elapsed since surgery. The morningto-evening variation was not consistent for the eyes studied at multiple times (1, 3.5, and II years). Compared with previous examinations, some eyes appeared to have stabilized, whereas some previously apparently stable eyes showed substantial morning-ta-evening changes at the 11-year visit.
A high degree of symmetry in diurnal fluctuation was observed between the two eyes of these patients, indicating that the amount of fluctuation in the second surgical eye
238
is likely to be similar to that in the first eye. This observation indicates that although the magnitude of diurnal fluctuation may not be predictable preoperatively in a given patient, those individuals who have this problem in their first eye are more likely to have it in the second eye, if they elect to have the surgery performed in the second eye.
Two ophthalmologists who had radial keratotomy have measured changes in their refraction from morning to evening during the first year after surgery.6,7 They measured their morning refractive errors sooner after awakening than we could in this study, and their data suggest that we may be underestimating the total amount of diurnal fluctuation by 1 D or more in some eyes.
Although there was not a statistically significant correlation between diurnal change in keratometry and refractive error, we believe the mechanism for these changes to be alteration in corneal topography. Because the keratometer examines such a small percentage of the anterior cornea, average keratometry power is a relatively insensitive measure of corneal refractive power changes after refractive surgery. The normal human cornea thickens 4% to 8% during sleep; most of this thickness increase disappears in the first hour or two after awakening,8,9 which could account for the marked change in refraction reported by Wyzinski and O'De11.6 MacRae and colleagues lO studied six patients in the early postoperative period after unilateral radial keratotomy surgery. Two weeks after the procedure, they found an average increase in myopia of 1.25 D from morning to evening in the surgical eye compared with 0.28 D in the nonsurgical eye. The mean central corneal thickness decreased 5.5% in the eye after radial keratotomy (0.599-0.569 mm) from morning to evening, whereas in the nonsurgical eye the decrease was 0.5% (0.558-0.555 mm). Intraocular pressure decreased by approximately 1.5 mmHg in both surgical and nonsurgical eyes. MacRae et allO suggested that the change in corneal thickness produced the measured changes in refractive error. In several other studies done at longer postoperative time intervals (months to years after surgery), however, no meaningful changes in central corneal thickness or intraocular pressure were measured that might explain the diurnal changes in refraction. I,2 Restoration of a completely intact epithelium over the radial incisions may limit the hydration changes of the cornea, but no studies have been able to determine the actual thicknesses within the incision scars, which might undergo hydration changes that are not measured by central pachometry. The absence of a correlation between change in central thickness and change in refractive error may suggest a role for mechanical factors, such as eyelid pressure, in changing the shape ofthe surgical cornea during waking hours. Variations in intraocular pressure also do not seem adequate to account for the corneal curvature and changes, and a correlation could not be shown in this study.
No patient characteristic measurable before surgery (e.g., patient age, magnitude of preoperative refractive error, intraocular pressure) allows us to predict which patients will have morning-to-evening refractive changes af-
McDonnell et al . Diurnal Changes at 11 Years in PERK
ter radial keratotomy. We did observe a high degree of symmetry in the morning-to-evening changes at 11 years after radial keratotomy in eyes of patients with bilateral surgery. Therefore, patient-specific variables, such as corneal biomechanical properties or the quality of the wound healing process, may determine the occurrence and severity of diurnal fluctuation after radial keratotomy.
The mechanism for the long-term refractive instability after radial keratotomy (as evidenced by increasing effect of surgery over years and persistence of diurnal fluctuation) is unknown. A major factor is probably the slow corneal wound healing process, with at least 3 to 5 years being required for the remodeling of unsutured keratotomy wounds. This has been demonstrated by finding epithelial plugs within wounds 5 years or longer after surgery.II-13 Remodeling of the stromal wounds also takes many years, as manifested by the presence of active fibroblasts seen histologically and biomicroscopic changes in wound morphology seen clinically.12 In addition, the tensile strength of the cornea is reduced after radial keratotomy as documented in laboratory studies and in reports of traumatic rupture of the cornea through radial keratotomy wounds. 14-16
Once the keratotomy wounds are remodeled-the epithelial plug ejected (ifit does not partly persist indefinitely in some incisions) and stromal remodeling completedthe tensile strength of the cornea may not return to normal. 17 It is not known whether the cornea after radial keratotomy ever assumes as stable a shape as that of the normal cornea once wound remodeling is completed. Careful studies of diurnal and long-term changes in refraction, keratometry, and visual acuity suggest that such stability has not been achieved even 11 years after surgery,3,18 a period of time by which most cellular wound repair activity has probably ceased.
References
1. Schanzlin DJ, 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.
2. Santos VR, Waring GO III, Lynn MJ, et al. Morning-toevening change in refraction, corneal curvature, and visual
acuity 2 to 4 years after radial keratotomy in the PERK Study. Ophthalmology 1988;95: 1487-93.
3, Waring GO III, Lynn MJ, McDonnell PJ, The PERK Study Group. Results of the Prospective Evaluation of Radial Keratotomy (PERK) Study ten years after surgery. Arch OphthalmoI1994;112:1298-1308.
4, Waring GO III, Moffitt SD, Gelender H, et al. Rationale for and design of the National Eye Institute Prospective Evaluation of Radial Keratotomy (PERK) Study. Ophthalmology 1983;90:40-58.
5, Waring GO III, Lynn MJ, Nizam A, et al. Results of the Prospective Evaluation of Radial Keratotomy (PERK) Study five years after surgery, Ophthalmology 1991 ;98: 1164-76.
6. Wyzinski P, O'Dell LW. Diurnal cycle of refraction after radial keratotomy, Ophthalmology 1987;94:120-4.
7, Richmond RD. Special report: radial keratotomy as seen through operated eyes. J Refract Surg 1987;3:22-7.
8, Mandell RB. Contact Lens Practice, 3rd ed. Springfield, IL: Charles C. Thomas, 1981.
9, Holden BA, Mertz GW, McNally JJ. Corneal swelling response to contact lenses worn under extended wear conditions. Invest Ophthalmol Vis Sci 1983;24:218-26,
10, MacRae S, Rich L, Phillips D, Bedrossian RH, Diurnal variation in vision after radial keratotomy. Am J Ophthalmol 1989;107:262-7.
11. Yamaguchi T, Tamaki K, Kaufman HE, et al. Histologic study of a pair of human corneas after anterior radial keratotomy. Am J Ophthalmol 1985;100:281-92.
12. Binder PS, Nayak SK, Deg JK, et al. An ultrastructural and histochemical study oflong-term wound healing after radial keratotomy. Am J Ophthalmol 1987;103:432-40.
13. Larson BC, Kremer FB, Eller A W, Bernardino VB Jr. Quantitated trauma following radial keratotomy in rabbits. Ophthalmology 1983;90:660-7.
14. Forstot SL, Damiano RE. Trauma after radial keratotomy. Ophthalmology 1988;95:833-5.
15. Bloom HR, Sands J, Schneider D. Corneal rupture from blunt trauma 22 months after radial keratotomy. Refract Corneal Surg 1990;6: 197 -9.
16. McDonnell PJ, Lean JS, Schanzlin DJ. Globe rupture from blunt trauma after hexagonal keratotomy [letter]. Am J Ophthalmol 1987; 103:241-2.
17. Bryant MR, Szerenyi K, Schmotzer H, McDonnell PJ. Corneal tensile strength in fully healed radial keratotomy wounds. Invest Ophthalmol Vis Sci 1994;35:3022-31.
18. Deitz MR, Sanders DR, Raanan MG, DeLuca M. Longterm (5- to 12-year) follow-up of metal-blade radial keratotomy procedures. Arch Ophthalmol 1994;112:614-20.
239
