- Email: [email protected]
A new concept for the correction of astigmatism: full-arc, depth-dependent astigmatic keratotomy
A New Concept for the Correction of Astigmatism: Full-Arc, Depth-Dependent Astigmatic Keratotomy Junsuke Akura, MD, PhD,1 Kazuki Matsuura, MD, PhD,1 Shiro Hatta, MD, PhD,2 Koji Otsuka, BA,1 Shuzo Kaneda, MD1 Objective: The purpose of this study is to introduce and evaluate a new concept in astigmatic keratotomy (AK) named full-arc, depth-dependent AK (FDAK). Design: Noncomparative interventional case series. Participants: FDAK was performed on a total of 37 eyes with regular astigmatism; of these, 16 eyes received FDAK alone, and 21 eyes received FDAK combined with cataract surgery. Methods: Corneal topography was used to divide the cornea into two discreet regions of “steep” and “flat.” Then, paired arcuate incisions, 90° in length, were placed along the full arc of the steep area. The level of astigmatic correction was controlled by varying the incision depth from 40% to 80% on the basis of a provisional nomogram developed by the authors. Main Outcome Measures: Keratometries, corneal topographies, and visual acuities were measured. Results: The FDAK alone group showed a significant improvement from a preoperative corneal astigmatism of 2.90 ⫾ 0.78 diopters (D) to a postoperative value of 0.89 ⫾ 0.52 D. The “combined” group also showed significant improvement from a preoperative corneal astigmatism of 2.97 ⫾ 1.01 D, to a postoperative value of 1.02 ⫾ 0.45 D. The deviation of achieved correction from attempted correction using vector analysis was between 1.37 D of undercorrection and 0.98 D of overcorrection, with 91.9% of cases within the range of ⫾1.0 D. Slight oblique change caused by axis deviation was observed in seven cases. Both uncorrected and corrected visual acuity showed statistically significant improvement. No serious complications were encountered. Conclusions: Controlling the level of correction by varying the incision depth allowed the surgeon to use long incisions (90° in length in regular astigmatism) covering the entire steep area, minimizing the undesirable changes induced by conventional deep and narrow incision AK and resulting in an ideal corneal sphericity after surgery. FDAK enabled the surgeon to accurately control the level of astigmatic correction with minimal risk of corneal perforation. Ophthalmology 2000;107:95–104 © 2000 by the American Academy of Ophthalmology. The conventional method of astigmatic keratotomy (AK) has dictated the use of a relatively deep incision fixed around 90% of corneal thickness. The length of incision has been the main factor manipulated in controlling the degree of astigmatic correction. However, it has been reported that in cases of AK in which the required incision length was relatively short (incision angle 30°– 60°), the prevalence of undesirable corneal changes seemed to increase, conse-
Originally received: September 14, 1998. Accepted: August 30, 1999. Manuscript no. 98485. 1 Department of Ophthalmology, Kushimoto Rehabilitation Center, Wakayama, Japan. 2 Department of Ophthalmology, Faculty of Medicine, Tottori University, Tottori, Japan. Presented in part as a video at the American Society of Cataract and Refractive Surgery (ASCRS) Meeting, Boston, Massachusetts, April 1997. Presented in part at the Japanese Convention of IOL and Refractive Surgery, Sendai, Japan, 1998. Reprint requests to Junsuke Akura, MD, Kushimoto Rehabilitation Center, 259-6 Kushimoto, Kushimoto-cho, Wakayama-ken, Japan. © 2000 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
quently the necessity of long-incision AK has been suggested.1,2 To induce an ideal corneal change and correct astigmatism more accurately by AK, we have devised a new method of AK, that uses a relatively long incision (90° in length in regular astigmatism) covering the full arc of the steep area and controls the degree of astigmatic correction by varying the incision depth. We have named this new concept of AK full-arc, depth-dependent AK (FDAK).
Patients and Methods Cases The FDAK cases reviewed in this study were performed on relatively symmetrical regular astigmatisms between June 1996 and June 1998. The 19 FDAK cases performed on asymmetrical astigmatism were eliminated from this study. The two patients who failed to follow up after the first month were eliminated from the study. We retrospectively reviewed a total of 37 eyes of 31 consecutive patients with almost regular astigmatism; of these, 16 eyes (13 patients) received FDAK alone, whereas 21 eyes (18 ISSN 0161-6420/00/$–see front matter PII S0161-6420(99)00021-4
95
Ophthalmology Volume 107, Number 1, January 2000 Table 1. Nomogram for FDAK Using Paired 90-degree Incisions with 7.5-mm Optical Zone Astigmatic Correction 5.0 4.0 3.0 2.0 1.0
D D D D D
Incision Depth (%) 80 70 60 50 40
FDAK ⫽ full-arc, depth-dependent astigmatic keratotomy. This nomogram is a version for regular astigmatism for older patients (older than 50). In regular astigmatism, FDAK always uses paired 90° incisions; intended incision depth is varied according to the magnitude of attempted astigmatic correction.
patients) received FDAK simultaneously with cataract surgery. All 37 cases were followed up for at least 6 months; 22 cases were followed up for 12 months. Among the “FDAK only” group, the average age was 64.2 years (range, 23– 85 years). Preoperative corneal astigmatism ranged from 1.5 to 3.75 D, with an average of 2.9 D. Fourteen eyes had against-the-rule (ATR) corneal astigmatism. One eye had with-the-rule (WTR) astigmatism. One eye had oblique astigmatism. Nine eyes had astigmatism caused by cataract surgery. Two eyes had astigmatism caused by glaucoma surgery. Five eyes had no operative history and were thought to be idiopathic. Among the “combined” operation group, the average age was 72.8 years, ranging from 34 to 95 years, with a preoperative corneal astigmatism average of 2.9 D (1.5– 6.0 D). The types of astigmatism were: 15 eyes with ATR astigmatism, 6 with WTR astigmatism. All these cases had no operation history and were considered to be idiopathic. For all the cataract surgeries, 3.0 to 3.5-mm “frown” incisions at oblique positions were used to induce the least amount of corneal astigmatism.3 The operations were performed by one surgeon at Kushimoto Rehabilitation Center in Japan. After discussing benefits and risks, informed consent was obtained before the surgery in all cases.
Planning of Surgery The FDAK surgery was planned according to the following protocol. First, the type of astigmatism was determined by analyzing the corneal topography. If the astigmatism was regular, paired incisions of 90° in length were planned along the steep axis. The incisions were arcuate, with an optical zone (OZ) of 7.5 mm. Next, the percentage of incision depth of local corneal thickness was decided according to our FDAK nomogram (Table 1), taking into account the magnitude of attempted astigmatic correction. Generally aiming for slight undercorrection after surgery, the degree of attempted correction was decided to be 70% to 90% of the keratometric values or delta K values from corneal topography.
Surgical Procedure The surgical procedures are as follows. To avoid axis deviation a reticule (Carl Zeiss, Oberkochen, Germany) was used preoperatively. The reticule allowed the surgeon to see “cross hairs” in the viewfinder of the slit lamp, enabling the surgeon to make a sketch of general “landmarks” (such as conjunctival vessels, pinguecula, and pigmentation) in relation to the vertical and horizontal axes. Just before operation, vertical and horizontal axes were marked by piokutanin pen on the cornea with the aid of the sketch. Then the
96
incision axes, decided by corneal topography, were marked. Next, under low light conditions the center of the entrance pupil4 was marked as the center for OZ. A Lindstrom Arcuate Incision Marker (with paired 90° arcuate lines, OZ 7 mm) covered with piokutanin was placed on the cornea so as to leave an ink imprint. Outside this imprint, dotted incision lines of 7.5-mm OZ were marked in ink. Ultrasonic pachymetry was performed at two locations on the incision line, on both sides of the center, until a consistent value was attained for each location. Then, the average of the two consistent values was calculated. This average was used as the local corneal thickness for determining incision depth using our nomogram. The same method was repeated to determine the incision depth of the second incision site as well. The incisions were made after the blade of the front-cutting diamond knife was adjusted to the planned incision depth. Cutting along the full length of the marked incision line often necessitated dividing the incision into two separate halves. The second half was made carefully so as to extend naturally from the first half. A depth gauge (Katena Products Inc, Denville, NJ) was used to make sure the entire incision was made at the intended depth. In places where the depth was too shallow, the knife was reinserted and the depth of incision was made constant.
Evaluation of Surgical Results and Postoperative Therapy Corrected and uncorrected visual acuity tests, keratometries, and topographies were conducted preoperatively and at 1 week, 1 month, 3 months, 6 months, and 1 year postoperatively. Keratometries were performed with Auto. Ref/Keratometer ARK-2000 (Nidek Inc, Gamagori, Japan). Topographies were taken by means of computer assisted videokeratography Eyesys 2000 (Eyesys Laboratories, Houston, TX). These examinations were performed by two nurses using standard charts. All data were entered into a networked computer database and subsequently analyzed. Analysis of astigmatism was performed by keratometry. Absolute cylinder without regard to axis and astigmatic change along the incision axis (using the vector analysis method of Holladay et al5) were analyzed. A paired Student’s t test was used for statistical analysis of astigmatism. As for the statistical analysis of visual acuity, the Wilcoxon singed-rank test was used after converting the visual acuity to a logarithmic scale. Ofloxacin (Talibit) and 0.1% fulorometron (0.1% Fulmetron) were administered four times daily for a period of approximately 1 month postoperatively.
Results Typical Cases of FDAK and Topographic Change Figures 1A–D show the typical cases of FDAK for correcting regular ATR astigmatism. Figures 1B, C are the cases of FDAK alone. Figures 1A and D are the cases of FDAK combined with cataract surgery. Each case had a different degree of preoperative keratometric astigmatism of 1.5 D ⫻ 94°, 2.5 D ⫻ 90°, 3.75 D ⫻ 84°, and 6.0 D ⫻ 90°, respectively. For these cases, the lengths of paired incisions were all 90°, but different depths of 40%, 50%, 60%, and 75% were used aiming for attempted corrections of 1.0 D, 2.0 D, 3.0 D, and 4.5 D, respectively. The 6-month postoperative keratometric values were 0.37 D ⫻ 135°, 0.5 ⫻ 120°, 0.75
Akura et al 䡠 Full-Arc Depth-Dependent Astigmatic Keratotomy
Figure 1. Examples of FDAK for almost regular astigmatism. A, Paired 90° incisions with 40% depth. B, Paired 90° incisions with 50% depth. (Fig 1 continues.)
D ⫻ 90°, and 1.5 D ⫻ 90°, respectively; vector analysis revealed astigmatic corrections of 1.49 D, 2.29 D, 3.02 D, and 4.5 D, respectively. For topographic change, every change map indicates that wide and effective flattening and steeping corresponding to the width and level of preoperative steep and flat area occurred. It is apparent that the magnitude of flattening and steepening increased as the incision was made deeper. All postoperative color maps show a marked improvement in corneal sphericity. In the cases of Figure 1A and D, 3.2-mm frown incisions were placed at an oblique location for cataract surgery. The topographic change thought to be induced by cataract incision cannot be observed. In the FDAK-alone surgery of Figure 1B, uncorrected visual acuity has improved from 20/50 to 20/25, and corrected visual acuity has improved from 20/30 to 20/16. Also in the FDAKalone case of Figure 1C, both uncorrected and corrected visual acuity improved from 20/50 to 20/20 and 20/25 to 20/15, respectively.
Astigmatic Change Figure 2 shows the postoperative course of mean absolute corneal astigmatism without regard to axis. In the group that underwent AK alone the absolute astigmatism averaged 2.90 ⫾ 0.78 D preoperatively and 0.89 ⫾ 0.52 D 6 months postoperatively. In the combined surgery group the absolute astigmatism averaged 2.97 ⫾ 1.01 D preoperatively and 1.01 ⫾ 0.45 D 6 months postoperatively. These differences were statistically significant (both P ⬍ 0.0001). Figure 3 shows the individual course of the residual astigmatism in the correcting axes using vector analysis. Twenty-two cases that were observed for more than 1 year were analyzed. A relatively strong fluctuation in astigmatism was observed in the early postoperative period; 15 eyes (68.2%) showed astigmatic regression of more than 0.5 D (maximum, 1.97 D) and 3 eyes (8.1%) showed astigmatic progression of more than 0.5 D (maximum, 1.03 D) between the first week and the first month. However, in general the fluctuation leveled off after the first month. Especially after the third month the astigmatism was stable; 21 cases (97.2%)
97
Ophthalmology Volume 107, Number 1, January 2000
Figure 1 (continued). C, Paired 90° incisions with 60% depth. D, Paired 90° incisions with 75% depth. These change maps indicate that wide and effective flattening and steepening occur. The degree of the flattening and steepening become stronger as the incisions become deeper.
showed astigmatic change of less than 0.5 D between 3 months and 12 months.
Incision Depth and Astigmatic Change (Attempted Correction and Achieved Correction) Figure 4 illustrates the relationship between intended incision depth and vector astigmatic change at 6 months. These results revealed that a relatively close correlation existed between the two. The attempted correction is decided by intended incision depth in the nomogram (Table 1), and achieved correction was calculated as vector astigmatic change, so Figure 4 also indicates the relationship between attempted and achieved correction at 6 months after surgery. The deviation of achieved correction from attempted correction ranged from an undercorrection of 1.37 D to an overcorrection of 0.98 D. All cases of younger patients (age 23– 44 years) showed relatively strong undercorrection, ranging from 0.75 D to 1.37 D (average, 1.01 D) of undercorrection. On the other hand, among the cases of older patients (age 57–95 years), patient age did not seem to affect results. When eliminating these five cases of
98
younger patients and estimating the results among the 32 cases for older patients, cases in which deviation was within the range of ⫾0.5 D and within the range of ⫾1.0 D were 25 eyes (87.5%) and 31 eyes (96.9%), respectively. In performing FDAK, the actual placement of incisions unintentionally fluctuated from an OZ of 7.0 mm to 8.3 mm. Postoperative analysis for older patients (age 57–95 years) revealed that the cases with small OZs of 7.0 to 7.2 mm (average, 7.04 mm) indicated a tendency of slight overcorrection (average, 0.25 D of overcorrection) and the cases with large OZs of 7.8 to 8.3 mm (average, 8.0 mm) indicated a tendency of slight undercorrection (average, 0.15 D of undercorrection). Cases with 7.3 to 7.7 mm OZs (average, 7.5 mm OZ) showed relatively favorable corrections (average, 0.09 D overcorrection).
Axial Deviation Table 2 illustrates the results of analyzing the deviation of postoperative axis from preoperative axis at 6 months. Thirty eyes (81.1%) showed on-axis change, whereas seven eyes (18.9%) showed oblique change. The seven cases which were considered to
Akura et al 䡠 Full-Arc Depth-Dependent Astigmatic Keratotomy
Figure 2. Postoperative course of mean (⫾SD) absolute keratometric astigmatism after FDAK (open squares) and FDAK combined with cataract surgery (closed squares).
be oblique change showed postoperative astigmatism of 1.0 D or less (averaged, 0.83 D).
Visual Acuity and Complications The preoperative and the 6-month postoperative visual acuities were compared in the “FDAK only” group. Assuming changes of two lines or more to be significant in uncorrected visual acuity, 14
Figure 4. Relationship between intended incision depth and vector astigmatic change at 6 months. This scattergram also indicates the relationship between attempted correction and achieved correction at 6 months. Closed circles (F), closed squares (■) and closed triangles (Œ) indicate the cases with incisions of 7.0 –7.2 mm OZ, 7.3–7.7 mm OZ, and 7.8 – 8.3 mm OZ respectively—for old patients (ages 57–95). Open squares (䊐) and open triangles (‚) indicate the cases with incisions of 7.3–7.7 mm OZ and 7.8 – 8.2 OZ respectively—for young patients (ages 23– 44); the patients’ ages are in parentheses.
eyes (87.5%) showed improvement, whereas 2 eyes (12.5%) exhibited no change. The geometric mean of uncorrected visual acuity rose from 20/91 preoperatively to 20/31 at 6 months postoperatively. For corrected visual acuity, 11 eyes (69%) were improved and 5 eyes (31%) exhibited no change. The geometric mean of corrected visual acuity was 20/31 preoperatively and was improved to 20/22 at 6 months postoperatively. Both uncorrected and corrected visual acuity showed statistically significant improvement (P ⫽ 0.0004, P ⫽ 0.0006, respectively). Complications such as intraoperative corneal perforation, postoperative infection, recurrent erosion, photophobia, and glare were not observed.
Discussion Ophthalmic surgery is currently undergoing a period of rapid transformation. Improvement in the quality of vision has become one of its major goals. FDAK is the result of reshaping the philosophy of conventional AK to better match today’s standards of the improvement in quality of vision. To explain the advantages of FDAK over conventional methods, we will first discuss the problems of conventional AK. Figure 3. Long-term individual course of residual astigmatism in the correcting axis using Holladay’s vector analysis. The value for residual astigmatism obtained by subtracting the value of vector astigmatic change from the preoperative astigmatic value. The correcting axis is defined as 90° away from the preoperative steep cylinder axis.
The Problems of Conventional AK According to the representative examples of conventional nomograms such as those of Thornton6 and Lindstrom,7 and
99
Ophthalmology Volume 107, Number 1, January 2000 Table 2. Axis Deviation ⬉0.5 D Degree of axis deviation No. of eyes ⫽ 37 Definition*
0° ⬃ 180° 13 (35.1%) On axis (complete correction)
>0.5 D 0° ⬃ 20° 160° ⬃ 180° 16 (43.2%) On axis (undercorrection)
21° ⬃ 69° 111° ⬃ 159° 7 (18.9%) Off axis (oblique change)
70° ⬃ 110° 1 (2.7%) On axis (overcorrection)
*Absolute (regardless of axis of astigmatism) postoperative cylinder ⬉0.5 D is defined to be an on-axis-complete correction here. Among absolute cylinder ⬎0.5 D; 160° ⬃ 180°, 0° ⬃ 20° axis deviations are defined as on-axis undercorrection, 70° ⬃ 110° axis deviations are defined as on-axis overcorrection, and 21° ⬃ 69°, 111° ⬃ 159° axis deviations are defined as off-axis oblique change.
others8 incision lengths of 30° to 60° are required to correct astigmatisms of 2.0 D to 4.0 D, which occur most frequently in clinical cases. However, incisions in the range of 30° to 60° are thought to be too short to cause ideal corneal sphericity postoperatively. The theoretical reasoning behind this belief will be explained using clinical cases and schematic diagrams.
Figure 5A, and B are the cases of conventional AK performed using paired short 45° incisions with 90% depth. Figure 5A is a case of on-axis incision (incision axis and preoperative steep axis the same). Preoperative and postoperative color maps show that the preoperative steep area appears to be split into two steep areas postoperatively (we call this split change). Figure 5B is a case of slight off-axis
Figure 5. Example of conventional short and deep incision AK A, Paired 45° incisions with 90% depth were placed on the steep axis. B, Paired 45° incisions with 90% depth were placed at the areas 15° shift from the steep axis. These change maps indicate that the areas of flattening are narrow. The steepening caused by coupling arises not only in the areas perpendicular to the incision site (perpendicular coupling) but also in areas adjacent to the flattening areas (adjacent coupling). As a result, on-axis incisions induced split change (A), 15° off-axis incisions induced oblique change (B) postoperatively.
100
Akura et al 䡠 Full-Arc Depth-Dependent Astigmatic Keratotomy incision (incision axis has shifted only 15° from the steep axis). Preoperative and postoperative color maps show that the strong steep area appears at the area 25° away from the preoperative steep area postoperatively (we call this oblique change). It should be noted on these change maps that, because incisions are short, the areas of flattening are narrow. Furthermore, steepenings caused by coupling9 were observed not only in areas perpendicular to the incision site but also in areas adjacent to the flattening. We coined the term “adjacent coupling” to describe the latter form of steepening. Figures 6c and d are diagrams that illustrate the representative characteristics of a change map for conventional AK using relatively short incisions of 30° to 60° at 90% depth. Figures 6a–f describe how these narrow flattenings and adjacent couplings can cause undesirable changes such as split change or oblique change postoperatively. In general, regular astigmatism can be divided into two regions of “steep” (indicated by hot colors) and “flat” (indicated by cold colors) by using the intermediate regions (indicated by green) as borders. When the incisions are placed on the steep axis and are shorter than the entire length of the steep areas (Fig 6a), the area within the incisions is flattened but steep areas remain that are not covered by the incisions. These remaining steep areas are increasingly steepened because of the adjacent coupling (Fig 6c) that occurs in these areas. As such, the preoperatively steep areas are split into two steep areas (Fig 6e). Also, even a slight axis deviation (Fig 6b) causes a severe oblique change (Fig 6f) because of the additional steepening caused by adjacent coupling (Fig 6d).
AK According to a New Paradigm To obtain an ideal corneal sphericity after surgery, the entire steep area must be flattened by the incision and the entire flat area must be steepened by coupling. To induce these changes, we believe that incisions in AK ideally must be made relatively long, so that the incisions cover the entire length of the preoperative steep area. Generally, for regular astigmatism, the steep area and flat area both cover a 90° area. We believed initially that incisions should be 90° in length in regular astigmatism. However, the flattening caused by such long incisions is too strong, resulting in overcorrection, we have developed a method of controlling the degree of correction by varying the incision depth between 40% and 80% of corneal thickness. The theoretical background of this method is based on the direct relationship (that we observed clinically) between incision depth and degree of astigmatic change in the range of 40% to 80% of corneal thickness. A diagram will be used to better explain these processes. Figures 1A–1D are the examples of FDAK performed with paired incisions 90° in length and 40% to 75% in depth. Figure 7c is the diagram of the characteristics of change caused by FDAK using paired incisions 90° in length and 40% to 80% in depth observed in numerous clinical cases. That is, when long incisions of 90° length are used, the flattening caused by the incision and the steepening caused by coupling both occur in broad regions perpen-
dicular to each other. The undesirable side effects caused by adjacent coupling (discussed earlier) do not occur. As a result, if long incisions (90 degrees in regular astigmatism) covering the entire steep areas are used (Fig 7a) and the incision depth is chosen to induce the appropriate degree of astigmatic change, the preoperative steep areas are totally flattened by the incisions, and the preoperative flat areas are totally steepened by effective coupling. As such, an ideal corneal sphericity can be obtained postoperatively (Fig 7e). It should also be noted that, because the incisions are long, the possibility of axis deviation is lessened. In this article, we restrict the report to the results of paired 90° incisions for correcting regular astigmatism. We are planning a full report on the methods and results for asymmetrical cases on another occasion. For cases of asymmetrical astigmatism, if incision length is varied according to the width of the steep area (Fig 7b), ideal corneal sphericity can be obtained postoperatively (Fig 7f).
Incision Depth and Astigmatic Change In the past, it has been believed that a shallow incision does not have enough effect for radial keratotomy and AK. However in terms of AK, the only studies we know of that have examined the relationship between incision depth and degree of astigmatic change were done by Seiler et al10 and Hanna et al.11 Seiler et al performed linear corneal T excisions with an excimer laser (Fig 8). Their graph illustrates that when incision depths of about 50% to 80% were used, the relationship between incision depth and the degree of astigmatic change are linearly correlated, but the level of astigmatic change was suddenly increased and highly fluctuated when incision depths of around 90% and greater were performed. We think that this result indicates that by varying the incision depths, the degree of astigmatic correction is controllable using incisions in the range of 50% to 80% in depth. On the other hand, the incisions with depths of 90% or more cause astigmatic changes, which fluctuated, and a predictable astigmatic correction cannot be performed. Hanna and colleagues describe the relationship between incision depth and astigmatic change in arcuate keratotomy using computer simulation. Their simulation demonstrates that deeper incisions produce a greater change; in general, incision depths of 30% or less produce small changes, incision depths greater than 70% produce large changes, and incision depths of 40% to 70% indicate a nearly linear relationship with incision depth (linear for incisions of 60 degrees in length, and a slightly exponential increase for incisions of 120 degrees in length). We think that this simulation also indicates that astigmatism can be controllable by varying the incision depth between 40% and 70%. Analysis of our clinical results revealed a relatively close correlation between degree of astigmatic change and incision depth in the range of 40% to 75% incision depth (Fig 4). This result confirmed that astigmatic correction can be controlled by varying the incision depth between 40% and 75%. Nonetheless, because of sparse data, further studies are needed before coming to a decisive conclusion. The method of slightly varying the incision depth for AK
101
Ophthalmology Volume 107, Number 1, January 2000
Figure 6. Schematic diagram of the preoperative and postoperative corneal shapes and changing maps characteristic of conventional short- and deepincision AK. If the incision is shorter than the full arc of the steep area (a, b), narrow flattening and adjacent coupling (c, d) induce split change (e) or oblique change (f) postoperatively.
has been reported by Gills12 and Guell et al.13 However, they use only two steps of incision depths (0.5 mm and 0.7 mm for Gills; 80% and 90% of corneal thickness for Guell) as a supporting means. As far as we know, FDAK is the first reported method in which astigmatic correction is controlled mainly by varying the incision depth.
Efficacy of FDAK To estimate the efficacy of surgery and the accuracy of the nomogram, the deviation between aimed and acquired astigmatic correction was studied (Fig 4). The deviation was observed to be between 1.37 D undercorrection and 0.98 D
Figure 7. Schematic diagram of the preoperative and postoperative corneal shapes and changing maps characteristic of FDAK. Because FDAK uses incisions covering the full arc of the steep areas (a, b) with varied depths to control the degree of astigmatic correction, the entire steep areas are effectively flattened by the incisions, and the entire flat areas are effectively steepened because of effective coupling (c, d), resulting in an ideal corneal sphericity postoperatively (e, f).
102
Akura et al 䡠 Full-Arc Depth-Dependent Astigmatic Keratotomy Using topographies, split change, which is often observed in short and deep incision AK, was rarely observed, and the center of the cornea is widely colored green in most topographies as shown in Figures 1A–D. Furthermore, corrected visual acuity and uncorrected visual acuity were significantly improved. These improvements are an indication that the cornea had attained an ideal sphericity and an improvement in the optical quality of the cornea had occurred. Another great advantage of FDAK to conventional AK is the minimal risk of corneal perforation because of its shallow incisions. At the time of FDAK combined with cataract surgery, the authors used 3.0 –3.5-mm frown incisions at an oblique position for getting an astigmatically neutral result in cataract surgery. We believe that all types of cataract incisions that are believed to be astigmatically neutral can be used in conjunction with FDAK. However, care should be taken so as not to overlap the FDAK incisions with cataract incisions, especially in clear corneal incisions, in which case astigmatic change may be greatly increased beyond the expected range (because of effects similar to a perforated AK). Figure 8. Relationship between incision depth and astigmatic change in human eyes treated with excimer laser using corneal T incisions. With the permission of Dr. Seiler, one of the erroneous data points has been corrected from the previously published print of this figure. (From Seiler et al. Am J Ophthalmol 1988;105:117–24.)
overcorrection, with 91.9% cases within the range of ⫾1.0 D. We think this is a reasonably high level of predictability/ accuracy. It should be noted that there was no case of strong overcorrection. All five cases of young patients (age 23– 44) revealed undercorrection of 0.75 D to 1.37 D (average, 1.01 D of undercorrection). Thus we believe a need exists for an additional nomogram to be constructed for younger patients (less than age 50). Although more clinical applications are needed to establish a nomogram for younger patients, our clinical experiences with younger patients suggest using incision depths that correspond to adding 1.0 D to the value of attempted correction of the current nomogram (Table 1). When we first began using FDAK, we speculated that the correction made by shallow incisions (40%– 80% depth) might not have a lasting effect. In fact, relatively strong regression was observed in several cases in the early postoperative period. However, these astigmatic fluctuations leveled off after the first month, and for the period between 3 months and 12 months almost every case showed stability of astigmatism (Fig 3). Because FDAK was a new procedure, we always aimed for slight undercorrection (70%–90% of complete correction). As a result, there was only one case of overcorrection, which indicated an approximately 90° shift in the astigmatic axis (Table 2). The cases of oblique change caused by axis deviation, a common problem in conventional AK, were seen in seven eyes (18.9%) also in our FDAK cases (Table 2). But these were not severe oblique changes as seen in Figure 4B. The postoperative astigmatisms of these cases were all 1.0 D or less. This indicates that these are very mild oblique changes.
Problems of FDAK The main difficulty in performing FDAK is in keeping the long arcuate incision neat and uniform. In the first few operations the incision became somewhat jagged or the incision strayed away from OZ marking line. Originally, the authors developed the nomogram for incisions of 7-mm OZ. But actual incisions fluctuated between an OZ of 7 mm and 8.3 mm. This is not only due to the technical difficulty in performing FDAK incisions but also to the surgeon’s habit of marking the OZ line slightly larger than the intended 7 mm and, similarly, making incisions slightly outside of the OZ marking line. As a result, the attempted correction value, which was obtained from the nomogram, corresponds well with surgically achieved correction in the cases with an OZ of approximately 7.5 mm. The cases with OZ of approximately 7 mm showed a tendency for overcorrection; and the cases with OZ of approximately 8 mm showed a tendency for undercorrection (Fig 4). This led the authors to change the OZ for the nomogram from the original OZ of 7.0 mm to an OZ of 7.5 mm. At present, we use an OZ of 7.5 mm for FDAK surgery. The surgical methods section of this article describes the method for performing 7.5-mm OZ FDAK surgery. In case the surgeon prefers a 7.0-mm or 8.0-mm OZ, we suggest modifying each of the values for attempted corrections on the nomogram by adding 0.25 D (for 7.0-mm OZ) or subtracting 0.15 D (for 8.0-mm OZ). As long as the incisions are performed by free hand, an occasional unintentional deviation from the OZ seems unavoidable. A more predictable surgery and a more precise evaluation of the surgical results depends on the instruments, such as the Hanna arcuate keratome (Moria, Antony, France), which creates accurate and reproducible incisions.14 Despite the success of our current FDAK nomogram, we feel the need to improve it further. We think that through
103
Ophthalmology Volume 107, Number 1, January 2000 many more clinical applications a more complete nomogram should be developed that also considers the factors of age. Similarly, variation in the size of OZ should be considered to obtain the most effective FDAK surgery. A multicenter study of FDAK is currently underway in Japan.
References 1. Ishii K, Miyata K. Corneal topography following astigmatic keratotomy after cataract surgery [in Japanese] [Eng. abstr.]. Jpn J Clin Ophthalmol 1993;47:1817–22. 2. Yoshitomi F, Hara Y. Refractive cataract surgery: superior small incision cataract surgery combined with paired astigmatic keratotomy [in Japanese]. Jpn J Clin Ophthalmol 1996; 50:124 –5. 3. Hatta S, Akura J. Control of astigmatism in cataract surgery: aiming for astigmatically neutral result and reducing preexisting astigmatism [in Japanese] [Eng. abstr.]. Jpn J Ophthal Surg 1994;7:451– 6. 4. Uozato H, Guyton DL. Centering corneal surgical procedures. Am J Ophthalmol 1987;103:264 –75. 5. Holladay JT, Cravy TV, Koch DD. Calculating the surgically induced refractive change following ocular surgery. J Cataract Refract Surg 1992;18:429 – 43.
104
6. Thornton SP. Astigmatic keratotomy: a review of basic concepts with case reports. J Cataract Refract Surg 1990; 16:30 –5. 7. Lindstrom RL. The surgical correction of astigmatism: a clinician’s perspective. Refract Corneal Surg 1990;6:441–54. 8. Price FW, Grene RB, Marks RG, et al. Astigmatism reduction clinical trial: a multicenter prospective evaluation of the predictability of arcuate keratotomy. Arch Opthalmol 1995;113: 277– 82. 9. Thornton SP. Background and theory of corneal relaxing incisions. In: Drummond AE, ed. Surgical Treatment of Astigmatism. Thoroface, NJ: SLACK Inc., 1994;1–9. 10. Seiler T, Bende T, Wollensak J, Trokel S. Excimer laser keratectomy for correction of astigmatism. Am J Ophthalmol 1988;105:117–24. 11. Hanna KD, Jouve FE, Waring GO III, Ciarlet PG. Computer simulation of arcuate keratotomy for astigmatism. In: Waring GO III, ed. Refractive Keratotomy for Myopia and Astigmatism. St. Louis: Mosby–year book, 1992;1249 –77. 12. Gills JP. Limbal relaxing incisions correct astigmatism predictably. Ocular Surg News 1997;15:14. 13. Guell JL, Manero F, Muller A. Transverse keratotomy to correct high corneal astigmatism after cataract surgery. J Cataract Refract Surg 1996;22:331– 6. 14. Hanna KD, Hayward JM, Hagen KB, et al. Keratotomy for astigmatism. Opthalmology 1987;103:264 –75.
Originally received: September 14, 1998. Accepted: August 30, 1999. Manuscript no. 98485. 1 Department of Ophthalmology, Kushimoto Rehabilitation Center, Wakayama, Japan. 2 Department of Ophthalmology, Faculty of Medicine, Tottori University, Tottori, Japan. Presented in part as a video at the American Society of Cataract and Refractive Surgery (ASCRS) Meeting, Boston, Massachusetts, April 1997. Presented in part at the Japanese Convention of IOL and Refractive Surgery, Sendai, Japan, 1998. Reprint requests to Junsuke Akura, MD, Kushimoto Rehabilitation Center, 259-6 Kushimoto, Kushimoto-cho, Wakayama-ken, Japan. © 2000 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
quently the necessity of long-incision AK has been suggested.1,2 To induce an ideal corneal change and correct astigmatism more accurately by AK, we have devised a new method of AK, that uses a relatively long incision (90° in length in regular astigmatism) covering the full arc of the steep area and controls the degree of astigmatic correction by varying the incision depth. We have named this new concept of AK full-arc, depth-dependent AK (FDAK).
Patients and Methods Cases The FDAK cases reviewed in this study were performed on relatively symmetrical regular astigmatisms between June 1996 and June 1998. The 19 FDAK cases performed on asymmetrical astigmatism were eliminated from this study. The two patients who failed to follow up after the first month were eliminated from the study. We retrospectively reviewed a total of 37 eyes of 31 consecutive patients with almost regular astigmatism; of these, 16 eyes (13 patients) received FDAK alone, whereas 21 eyes (18 ISSN 0161-6420/00/$–see front matter PII S0161-6420(99)00021-4
95
Ophthalmology Volume 107, Number 1, January 2000 Table 1. Nomogram for FDAK Using Paired 90-degree Incisions with 7.5-mm Optical Zone Astigmatic Correction 5.0 4.0 3.0 2.0 1.0
D D D D D
Incision Depth (%) 80 70 60 50 40
FDAK ⫽ full-arc, depth-dependent astigmatic keratotomy. This nomogram is a version for regular astigmatism for older patients (older than 50). In regular astigmatism, FDAK always uses paired 90° incisions; intended incision depth is varied according to the magnitude of attempted astigmatic correction.
patients) received FDAK simultaneously with cataract surgery. All 37 cases were followed up for at least 6 months; 22 cases were followed up for 12 months. Among the “FDAK only” group, the average age was 64.2 years (range, 23– 85 years). Preoperative corneal astigmatism ranged from 1.5 to 3.75 D, with an average of 2.9 D. Fourteen eyes had against-the-rule (ATR) corneal astigmatism. One eye had with-the-rule (WTR) astigmatism. One eye had oblique astigmatism. Nine eyes had astigmatism caused by cataract surgery. Two eyes had astigmatism caused by glaucoma surgery. Five eyes had no operative history and were thought to be idiopathic. Among the “combined” operation group, the average age was 72.8 years, ranging from 34 to 95 years, with a preoperative corneal astigmatism average of 2.9 D (1.5– 6.0 D). The types of astigmatism were: 15 eyes with ATR astigmatism, 6 with WTR astigmatism. All these cases had no operation history and were considered to be idiopathic. For all the cataract surgeries, 3.0 to 3.5-mm “frown” incisions at oblique positions were used to induce the least amount of corneal astigmatism.3 The operations were performed by one surgeon at Kushimoto Rehabilitation Center in Japan. After discussing benefits and risks, informed consent was obtained before the surgery in all cases.
Planning of Surgery The FDAK surgery was planned according to the following protocol. First, the type of astigmatism was determined by analyzing the corneal topography. If the astigmatism was regular, paired incisions of 90° in length were planned along the steep axis. The incisions were arcuate, with an optical zone (OZ) of 7.5 mm. Next, the percentage of incision depth of local corneal thickness was decided according to our FDAK nomogram (Table 1), taking into account the magnitude of attempted astigmatic correction. Generally aiming for slight undercorrection after surgery, the degree of attempted correction was decided to be 70% to 90% of the keratometric values or delta K values from corneal topography.
Surgical Procedure The surgical procedures are as follows. To avoid axis deviation a reticule (Carl Zeiss, Oberkochen, Germany) was used preoperatively. The reticule allowed the surgeon to see “cross hairs” in the viewfinder of the slit lamp, enabling the surgeon to make a sketch of general “landmarks” (such as conjunctival vessels, pinguecula, and pigmentation) in relation to the vertical and horizontal axes. Just before operation, vertical and horizontal axes were marked by piokutanin pen on the cornea with the aid of the sketch. Then the
96
incision axes, decided by corneal topography, were marked. Next, under low light conditions the center of the entrance pupil4 was marked as the center for OZ. A Lindstrom Arcuate Incision Marker (with paired 90° arcuate lines, OZ 7 mm) covered with piokutanin was placed on the cornea so as to leave an ink imprint. Outside this imprint, dotted incision lines of 7.5-mm OZ were marked in ink. Ultrasonic pachymetry was performed at two locations on the incision line, on both sides of the center, until a consistent value was attained for each location. Then, the average of the two consistent values was calculated. This average was used as the local corneal thickness for determining incision depth using our nomogram. The same method was repeated to determine the incision depth of the second incision site as well. The incisions were made after the blade of the front-cutting diamond knife was adjusted to the planned incision depth. Cutting along the full length of the marked incision line often necessitated dividing the incision into two separate halves. The second half was made carefully so as to extend naturally from the first half. A depth gauge (Katena Products Inc, Denville, NJ) was used to make sure the entire incision was made at the intended depth. In places where the depth was too shallow, the knife was reinserted and the depth of incision was made constant.
Evaluation of Surgical Results and Postoperative Therapy Corrected and uncorrected visual acuity tests, keratometries, and topographies were conducted preoperatively and at 1 week, 1 month, 3 months, 6 months, and 1 year postoperatively. Keratometries were performed with Auto. Ref/Keratometer ARK-2000 (Nidek Inc, Gamagori, Japan). Topographies were taken by means of computer assisted videokeratography Eyesys 2000 (Eyesys Laboratories, Houston, TX). These examinations were performed by two nurses using standard charts. All data were entered into a networked computer database and subsequently analyzed. Analysis of astigmatism was performed by keratometry. Absolute cylinder without regard to axis and astigmatic change along the incision axis (using the vector analysis method of Holladay et al5) were analyzed. A paired Student’s t test was used for statistical analysis of astigmatism. As for the statistical analysis of visual acuity, the Wilcoxon singed-rank test was used after converting the visual acuity to a logarithmic scale. Ofloxacin (Talibit) and 0.1% fulorometron (0.1% Fulmetron) were administered four times daily for a period of approximately 1 month postoperatively.
Results Typical Cases of FDAK and Topographic Change Figures 1A–D show the typical cases of FDAK for correcting regular ATR astigmatism. Figures 1B, C are the cases of FDAK alone. Figures 1A and D are the cases of FDAK combined with cataract surgery. Each case had a different degree of preoperative keratometric astigmatism of 1.5 D ⫻ 94°, 2.5 D ⫻ 90°, 3.75 D ⫻ 84°, and 6.0 D ⫻ 90°, respectively. For these cases, the lengths of paired incisions were all 90°, but different depths of 40%, 50%, 60%, and 75% were used aiming for attempted corrections of 1.0 D, 2.0 D, 3.0 D, and 4.5 D, respectively. The 6-month postoperative keratometric values were 0.37 D ⫻ 135°, 0.5 ⫻ 120°, 0.75
Akura et al 䡠 Full-Arc Depth-Dependent Astigmatic Keratotomy
Figure 1. Examples of FDAK for almost regular astigmatism. A, Paired 90° incisions with 40% depth. B, Paired 90° incisions with 50% depth. (Fig 1 continues.)
D ⫻ 90°, and 1.5 D ⫻ 90°, respectively; vector analysis revealed astigmatic corrections of 1.49 D, 2.29 D, 3.02 D, and 4.5 D, respectively. For topographic change, every change map indicates that wide and effective flattening and steeping corresponding to the width and level of preoperative steep and flat area occurred. It is apparent that the magnitude of flattening and steepening increased as the incision was made deeper. All postoperative color maps show a marked improvement in corneal sphericity. In the cases of Figure 1A and D, 3.2-mm frown incisions were placed at an oblique location for cataract surgery. The topographic change thought to be induced by cataract incision cannot be observed. In the FDAK-alone surgery of Figure 1B, uncorrected visual acuity has improved from 20/50 to 20/25, and corrected visual acuity has improved from 20/30 to 20/16. Also in the FDAKalone case of Figure 1C, both uncorrected and corrected visual acuity improved from 20/50 to 20/20 and 20/25 to 20/15, respectively.
Astigmatic Change Figure 2 shows the postoperative course of mean absolute corneal astigmatism without regard to axis. In the group that underwent AK alone the absolute astigmatism averaged 2.90 ⫾ 0.78 D preoperatively and 0.89 ⫾ 0.52 D 6 months postoperatively. In the combined surgery group the absolute astigmatism averaged 2.97 ⫾ 1.01 D preoperatively and 1.01 ⫾ 0.45 D 6 months postoperatively. These differences were statistically significant (both P ⬍ 0.0001). Figure 3 shows the individual course of the residual astigmatism in the correcting axes using vector analysis. Twenty-two cases that were observed for more than 1 year were analyzed. A relatively strong fluctuation in astigmatism was observed in the early postoperative period; 15 eyes (68.2%) showed astigmatic regression of more than 0.5 D (maximum, 1.97 D) and 3 eyes (8.1%) showed astigmatic progression of more than 0.5 D (maximum, 1.03 D) between the first week and the first month. However, in general the fluctuation leveled off after the first month. Especially after the third month the astigmatism was stable; 21 cases (97.2%)
97
Ophthalmology Volume 107, Number 1, January 2000
Figure 1 (continued). C, Paired 90° incisions with 60% depth. D, Paired 90° incisions with 75% depth. These change maps indicate that wide and effective flattening and steepening occur. The degree of the flattening and steepening become stronger as the incisions become deeper.
showed astigmatic change of less than 0.5 D between 3 months and 12 months.
Incision Depth and Astigmatic Change (Attempted Correction and Achieved Correction) Figure 4 illustrates the relationship between intended incision depth and vector astigmatic change at 6 months. These results revealed that a relatively close correlation existed between the two. The attempted correction is decided by intended incision depth in the nomogram (Table 1), and achieved correction was calculated as vector astigmatic change, so Figure 4 also indicates the relationship between attempted and achieved correction at 6 months after surgery. The deviation of achieved correction from attempted correction ranged from an undercorrection of 1.37 D to an overcorrection of 0.98 D. All cases of younger patients (age 23– 44 years) showed relatively strong undercorrection, ranging from 0.75 D to 1.37 D (average, 1.01 D) of undercorrection. On the other hand, among the cases of older patients (age 57–95 years), patient age did not seem to affect results. When eliminating these five cases of
98
younger patients and estimating the results among the 32 cases for older patients, cases in which deviation was within the range of ⫾0.5 D and within the range of ⫾1.0 D were 25 eyes (87.5%) and 31 eyes (96.9%), respectively. In performing FDAK, the actual placement of incisions unintentionally fluctuated from an OZ of 7.0 mm to 8.3 mm. Postoperative analysis for older patients (age 57–95 years) revealed that the cases with small OZs of 7.0 to 7.2 mm (average, 7.04 mm) indicated a tendency of slight overcorrection (average, 0.25 D of overcorrection) and the cases with large OZs of 7.8 to 8.3 mm (average, 8.0 mm) indicated a tendency of slight undercorrection (average, 0.15 D of undercorrection). Cases with 7.3 to 7.7 mm OZs (average, 7.5 mm OZ) showed relatively favorable corrections (average, 0.09 D overcorrection).
Axial Deviation Table 2 illustrates the results of analyzing the deviation of postoperative axis from preoperative axis at 6 months. Thirty eyes (81.1%) showed on-axis change, whereas seven eyes (18.9%) showed oblique change. The seven cases which were considered to
Akura et al 䡠 Full-Arc Depth-Dependent Astigmatic Keratotomy
Figure 2. Postoperative course of mean (⫾SD) absolute keratometric astigmatism after FDAK (open squares) and FDAK combined with cataract surgery (closed squares).
be oblique change showed postoperative astigmatism of 1.0 D or less (averaged, 0.83 D).
Visual Acuity and Complications The preoperative and the 6-month postoperative visual acuities were compared in the “FDAK only” group. Assuming changes of two lines or more to be significant in uncorrected visual acuity, 14
Figure 4. Relationship between intended incision depth and vector astigmatic change at 6 months. This scattergram also indicates the relationship between attempted correction and achieved correction at 6 months. Closed circles (F), closed squares (■) and closed triangles (Œ) indicate the cases with incisions of 7.0 –7.2 mm OZ, 7.3–7.7 mm OZ, and 7.8 – 8.3 mm OZ respectively—for old patients (ages 57–95). Open squares (䊐) and open triangles (‚) indicate the cases with incisions of 7.3–7.7 mm OZ and 7.8 – 8.2 OZ respectively—for young patients (ages 23– 44); the patients’ ages are in parentheses.
eyes (87.5%) showed improvement, whereas 2 eyes (12.5%) exhibited no change. The geometric mean of uncorrected visual acuity rose from 20/91 preoperatively to 20/31 at 6 months postoperatively. For corrected visual acuity, 11 eyes (69%) were improved and 5 eyes (31%) exhibited no change. The geometric mean of corrected visual acuity was 20/31 preoperatively and was improved to 20/22 at 6 months postoperatively. Both uncorrected and corrected visual acuity showed statistically significant improvement (P ⫽ 0.0004, P ⫽ 0.0006, respectively). Complications such as intraoperative corneal perforation, postoperative infection, recurrent erosion, photophobia, and glare were not observed.
Discussion Ophthalmic surgery is currently undergoing a period of rapid transformation. Improvement in the quality of vision has become one of its major goals. FDAK is the result of reshaping the philosophy of conventional AK to better match today’s standards of the improvement in quality of vision. To explain the advantages of FDAK over conventional methods, we will first discuss the problems of conventional AK. Figure 3. Long-term individual course of residual astigmatism in the correcting axis using Holladay’s vector analysis. The value for residual astigmatism obtained by subtracting the value of vector astigmatic change from the preoperative astigmatic value. The correcting axis is defined as 90° away from the preoperative steep cylinder axis.
The Problems of Conventional AK According to the representative examples of conventional nomograms such as those of Thornton6 and Lindstrom,7 and
99
Ophthalmology Volume 107, Number 1, January 2000 Table 2. Axis Deviation ⬉0.5 D Degree of axis deviation No. of eyes ⫽ 37 Definition*
0° ⬃ 180° 13 (35.1%) On axis (complete correction)
>0.5 D 0° ⬃ 20° 160° ⬃ 180° 16 (43.2%) On axis (undercorrection)
21° ⬃ 69° 111° ⬃ 159° 7 (18.9%) Off axis (oblique change)
70° ⬃ 110° 1 (2.7%) On axis (overcorrection)
*Absolute (regardless of axis of astigmatism) postoperative cylinder ⬉0.5 D is defined to be an on-axis-complete correction here. Among absolute cylinder ⬎0.5 D; 160° ⬃ 180°, 0° ⬃ 20° axis deviations are defined as on-axis undercorrection, 70° ⬃ 110° axis deviations are defined as on-axis overcorrection, and 21° ⬃ 69°, 111° ⬃ 159° axis deviations are defined as off-axis oblique change.
others8 incision lengths of 30° to 60° are required to correct astigmatisms of 2.0 D to 4.0 D, which occur most frequently in clinical cases. However, incisions in the range of 30° to 60° are thought to be too short to cause ideal corneal sphericity postoperatively. The theoretical reasoning behind this belief will be explained using clinical cases and schematic diagrams.
Figure 5A, and B are the cases of conventional AK performed using paired short 45° incisions with 90% depth. Figure 5A is a case of on-axis incision (incision axis and preoperative steep axis the same). Preoperative and postoperative color maps show that the preoperative steep area appears to be split into two steep areas postoperatively (we call this split change). Figure 5B is a case of slight off-axis
Figure 5. Example of conventional short and deep incision AK A, Paired 45° incisions with 90% depth were placed on the steep axis. B, Paired 45° incisions with 90% depth were placed at the areas 15° shift from the steep axis. These change maps indicate that the areas of flattening are narrow. The steepening caused by coupling arises not only in the areas perpendicular to the incision site (perpendicular coupling) but also in areas adjacent to the flattening areas (adjacent coupling). As a result, on-axis incisions induced split change (A), 15° off-axis incisions induced oblique change (B) postoperatively.
100
Akura et al 䡠 Full-Arc Depth-Dependent Astigmatic Keratotomy incision (incision axis has shifted only 15° from the steep axis). Preoperative and postoperative color maps show that the strong steep area appears at the area 25° away from the preoperative steep area postoperatively (we call this oblique change). It should be noted on these change maps that, because incisions are short, the areas of flattening are narrow. Furthermore, steepenings caused by coupling9 were observed not only in areas perpendicular to the incision site but also in areas adjacent to the flattening. We coined the term “adjacent coupling” to describe the latter form of steepening. Figures 6c and d are diagrams that illustrate the representative characteristics of a change map for conventional AK using relatively short incisions of 30° to 60° at 90% depth. Figures 6a–f describe how these narrow flattenings and adjacent couplings can cause undesirable changes such as split change or oblique change postoperatively. In general, regular astigmatism can be divided into two regions of “steep” (indicated by hot colors) and “flat” (indicated by cold colors) by using the intermediate regions (indicated by green) as borders. When the incisions are placed on the steep axis and are shorter than the entire length of the steep areas (Fig 6a), the area within the incisions is flattened but steep areas remain that are not covered by the incisions. These remaining steep areas are increasingly steepened because of the adjacent coupling (Fig 6c) that occurs in these areas. As such, the preoperatively steep areas are split into two steep areas (Fig 6e). Also, even a slight axis deviation (Fig 6b) causes a severe oblique change (Fig 6f) because of the additional steepening caused by adjacent coupling (Fig 6d).
AK According to a New Paradigm To obtain an ideal corneal sphericity after surgery, the entire steep area must be flattened by the incision and the entire flat area must be steepened by coupling. To induce these changes, we believe that incisions in AK ideally must be made relatively long, so that the incisions cover the entire length of the preoperative steep area. Generally, for regular astigmatism, the steep area and flat area both cover a 90° area. We believed initially that incisions should be 90° in length in regular astigmatism. However, the flattening caused by such long incisions is too strong, resulting in overcorrection, we have developed a method of controlling the degree of correction by varying the incision depth between 40% and 80% of corneal thickness. The theoretical background of this method is based on the direct relationship (that we observed clinically) between incision depth and degree of astigmatic change in the range of 40% to 80% of corneal thickness. A diagram will be used to better explain these processes. Figures 1A–1D are the examples of FDAK performed with paired incisions 90° in length and 40% to 75% in depth. Figure 7c is the diagram of the characteristics of change caused by FDAK using paired incisions 90° in length and 40% to 80% in depth observed in numerous clinical cases. That is, when long incisions of 90° length are used, the flattening caused by the incision and the steepening caused by coupling both occur in broad regions perpen-
dicular to each other. The undesirable side effects caused by adjacent coupling (discussed earlier) do not occur. As a result, if long incisions (90 degrees in regular astigmatism) covering the entire steep areas are used (Fig 7a) and the incision depth is chosen to induce the appropriate degree of astigmatic change, the preoperative steep areas are totally flattened by the incisions, and the preoperative flat areas are totally steepened by effective coupling. As such, an ideal corneal sphericity can be obtained postoperatively (Fig 7e). It should also be noted that, because the incisions are long, the possibility of axis deviation is lessened. In this article, we restrict the report to the results of paired 90° incisions for correcting regular astigmatism. We are planning a full report on the methods and results for asymmetrical cases on another occasion. For cases of asymmetrical astigmatism, if incision length is varied according to the width of the steep area (Fig 7b), ideal corneal sphericity can be obtained postoperatively (Fig 7f).
Incision Depth and Astigmatic Change In the past, it has been believed that a shallow incision does not have enough effect for radial keratotomy and AK. However in terms of AK, the only studies we know of that have examined the relationship between incision depth and degree of astigmatic change were done by Seiler et al10 and Hanna et al.11 Seiler et al performed linear corneal T excisions with an excimer laser (Fig 8). Their graph illustrates that when incision depths of about 50% to 80% were used, the relationship between incision depth and the degree of astigmatic change are linearly correlated, but the level of astigmatic change was suddenly increased and highly fluctuated when incision depths of around 90% and greater were performed. We think that this result indicates that by varying the incision depths, the degree of astigmatic correction is controllable using incisions in the range of 50% to 80% in depth. On the other hand, the incisions with depths of 90% or more cause astigmatic changes, which fluctuated, and a predictable astigmatic correction cannot be performed. Hanna and colleagues describe the relationship between incision depth and astigmatic change in arcuate keratotomy using computer simulation. Their simulation demonstrates that deeper incisions produce a greater change; in general, incision depths of 30% or less produce small changes, incision depths greater than 70% produce large changes, and incision depths of 40% to 70% indicate a nearly linear relationship with incision depth (linear for incisions of 60 degrees in length, and a slightly exponential increase for incisions of 120 degrees in length). We think that this simulation also indicates that astigmatism can be controllable by varying the incision depth between 40% and 70%. Analysis of our clinical results revealed a relatively close correlation between degree of astigmatic change and incision depth in the range of 40% to 75% incision depth (Fig 4). This result confirmed that astigmatic correction can be controlled by varying the incision depth between 40% and 75%. Nonetheless, because of sparse data, further studies are needed before coming to a decisive conclusion. The method of slightly varying the incision depth for AK
101
Ophthalmology Volume 107, Number 1, January 2000
Figure 6. Schematic diagram of the preoperative and postoperative corneal shapes and changing maps characteristic of conventional short- and deepincision AK. If the incision is shorter than the full arc of the steep area (a, b), narrow flattening and adjacent coupling (c, d) induce split change (e) or oblique change (f) postoperatively.
has been reported by Gills12 and Guell et al.13 However, they use only two steps of incision depths (0.5 mm and 0.7 mm for Gills; 80% and 90% of corneal thickness for Guell) as a supporting means. As far as we know, FDAK is the first reported method in which astigmatic correction is controlled mainly by varying the incision depth.
Efficacy of FDAK To estimate the efficacy of surgery and the accuracy of the nomogram, the deviation between aimed and acquired astigmatic correction was studied (Fig 4). The deviation was observed to be between 1.37 D undercorrection and 0.98 D
Figure 7. Schematic diagram of the preoperative and postoperative corneal shapes and changing maps characteristic of FDAK. Because FDAK uses incisions covering the full arc of the steep areas (a, b) with varied depths to control the degree of astigmatic correction, the entire steep areas are effectively flattened by the incisions, and the entire flat areas are effectively steepened because of effective coupling (c, d), resulting in an ideal corneal sphericity postoperatively (e, f).
102
Akura et al 䡠 Full-Arc Depth-Dependent Astigmatic Keratotomy Using topographies, split change, which is often observed in short and deep incision AK, was rarely observed, and the center of the cornea is widely colored green in most topographies as shown in Figures 1A–D. Furthermore, corrected visual acuity and uncorrected visual acuity were significantly improved. These improvements are an indication that the cornea had attained an ideal sphericity and an improvement in the optical quality of the cornea had occurred. Another great advantage of FDAK to conventional AK is the minimal risk of corneal perforation because of its shallow incisions. At the time of FDAK combined with cataract surgery, the authors used 3.0 –3.5-mm frown incisions at an oblique position for getting an astigmatically neutral result in cataract surgery. We believe that all types of cataract incisions that are believed to be astigmatically neutral can be used in conjunction with FDAK. However, care should be taken so as not to overlap the FDAK incisions with cataract incisions, especially in clear corneal incisions, in which case astigmatic change may be greatly increased beyond the expected range (because of effects similar to a perforated AK). Figure 8. Relationship between incision depth and astigmatic change in human eyes treated with excimer laser using corneal T incisions. With the permission of Dr. Seiler, one of the erroneous data points has been corrected from the previously published print of this figure. (From Seiler et al. Am J Ophthalmol 1988;105:117–24.)
overcorrection, with 91.9% cases within the range of ⫾1.0 D. We think this is a reasonably high level of predictability/ accuracy. It should be noted that there was no case of strong overcorrection. All five cases of young patients (age 23– 44) revealed undercorrection of 0.75 D to 1.37 D (average, 1.01 D of undercorrection). Thus we believe a need exists for an additional nomogram to be constructed for younger patients (less than age 50). Although more clinical applications are needed to establish a nomogram for younger patients, our clinical experiences with younger patients suggest using incision depths that correspond to adding 1.0 D to the value of attempted correction of the current nomogram (Table 1). When we first began using FDAK, we speculated that the correction made by shallow incisions (40%– 80% depth) might not have a lasting effect. In fact, relatively strong regression was observed in several cases in the early postoperative period. However, these astigmatic fluctuations leveled off after the first month, and for the period between 3 months and 12 months almost every case showed stability of astigmatism (Fig 3). Because FDAK was a new procedure, we always aimed for slight undercorrection (70%–90% of complete correction). As a result, there was only one case of overcorrection, which indicated an approximately 90° shift in the astigmatic axis (Table 2). The cases of oblique change caused by axis deviation, a common problem in conventional AK, were seen in seven eyes (18.9%) also in our FDAK cases (Table 2). But these were not severe oblique changes as seen in Figure 4B. The postoperative astigmatisms of these cases were all 1.0 D or less. This indicates that these are very mild oblique changes.
Problems of FDAK The main difficulty in performing FDAK is in keeping the long arcuate incision neat and uniform. In the first few operations the incision became somewhat jagged or the incision strayed away from OZ marking line. Originally, the authors developed the nomogram for incisions of 7-mm OZ. But actual incisions fluctuated between an OZ of 7 mm and 8.3 mm. This is not only due to the technical difficulty in performing FDAK incisions but also to the surgeon’s habit of marking the OZ line slightly larger than the intended 7 mm and, similarly, making incisions slightly outside of the OZ marking line. As a result, the attempted correction value, which was obtained from the nomogram, corresponds well with surgically achieved correction in the cases with an OZ of approximately 7.5 mm. The cases with OZ of approximately 7 mm showed a tendency for overcorrection; and the cases with OZ of approximately 8 mm showed a tendency for undercorrection (Fig 4). This led the authors to change the OZ for the nomogram from the original OZ of 7.0 mm to an OZ of 7.5 mm. At present, we use an OZ of 7.5 mm for FDAK surgery. The surgical methods section of this article describes the method for performing 7.5-mm OZ FDAK surgery. In case the surgeon prefers a 7.0-mm or 8.0-mm OZ, we suggest modifying each of the values for attempted corrections on the nomogram by adding 0.25 D (for 7.0-mm OZ) or subtracting 0.15 D (for 8.0-mm OZ). As long as the incisions are performed by free hand, an occasional unintentional deviation from the OZ seems unavoidable. A more predictable surgery and a more precise evaluation of the surgical results depends on the instruments, such as the Hanna arcuate keratome (Moria, Antony, France), which creates accurate and reproducible incisions.14 Despite the success of our current FDAK nomogram, we feel the need to improve it further. We think that through
103
Ophthalmology Volume 107, Number 1, January 2000 many more clinical applications a more complete nomogram should be developed that also considers the factors of age. Similarly, variation in the size of OZ should be considered to obtain the most effective FDAK surgery. A multicenter study of FDAK is currently underway in Japan.
References 1. Ishii K, Miyata K. Corneal topography following astigmatic keratotomy after cataract surgery [in Japanese] [Eng. abstr.]. Jpn J Clin Ophthalmol 1993;47:1817–22. 2. Yoshitomi F, Hara Y. Refractive cataract surgery: superior small incision cataract surgery combined with paired astigmatic keratotomy [in Japanese]. Jpn J Clin Ophthalmol 1996; 50:124 –5. 3. Hatta S, Akura J. Control of astigmatism in cataract surgery: aiming for astigmatically neutral result and reducing preexisting astigmatism [in Japanese] [Eng. abstr.]. Jpn J Ophthal Surg 1994;7:451– 6. 4. Uozato H, Guyton DL. Centering corneal surgical procedures. Am J Ophthalmol 1987;103:264 –75. 5. Holladay JT, Cravy TV, Koch DD. Calculating the surgically induced refractive change following ocular surgery. J Cataract Refract Surg 1992;18:429 – 43.
104
6. Thornton SP. Astigmatic keratotomy: a review of basic concepts with case reports. J Cataract Refract Surg 1990; 16:30 –5. 7. Lindstrom RL. The surgical correction of astigmatism: a clinician’s perspective. Refract Corneal Surg 1990;6:441–54. 8. Price FW, Grene RB, Marks RG, et al. Astigmatism reduction clinical trial: a multicenter prospective evaluation of the predictability of arcuate keratotomy. Arch Opthalmol 1995;113: 277– 82. 9. Thornton SP. Background and theory of corneal relaxing incisions. In: Drummond AE, ed. Surgical Treatment of Astigmatism. Thoroface, NJ: SLACK Inc., 1994;1–9. 10. Seiler T, Bende T, Wollensak J, Trokel S. Excimer laser keratectomy for correction of astigmatism. Am J Ophthalmol 1988;105:117–24. 11. Hanna KD, Jouve FE, Waring GO III, Ciarlet PG. Computer simulation of arcuate keratotomy for astigmatism. In: Waring GO III, ed. Refractive Keratotomy for Myopia and Astigmatism. St. Louis: Mosby–year book, 1992;1249 –77. 12. Gills JP. Limbal relaxing incisions correct astigmatism predictably. Ocular Surg News 1997;15:14. 13. Guell JL, Manero F, Muller A. Transverse keratotomy to correct high corneal astigmatism after cataract surgery. J Cataract Refract Surg 1996;22:331– 6. 14. Hanna KD, Hayward JM, Hagen KB, et al. Keratotomy for astigmatism. Opthalmology 1987;103:264 –75.