Recent diagnostic and surgical ophthalmic instruments, especially in the field of refractive surgery, rely on information such as the line of sight, visual axis, fixative lines, or vertex normal. Even small decentrations of 0.5 mm of the ablation zone in laser in situ keratomileusis (LASIK) or photorefractive keratectomy can cause irregular astigmatism and result in dramatic reduction of best spectacle-corrected visual acuity.1
Many excimer lasers are now equipped with eye-tracking systems.1–5 Most of these systems consider the pupil margin using an infrared camera and follow the center of the entrance pupil. This principal of centration6 is based on previous reports showing that photoreceptors orient themselves toward the center of the eccentric pupil and that light passing through the center of the eye is more effective in stimulating the photoreceptors than light passing through the peripheral pupil.7–8
Mandell9,10 reported that eyes with a larger angle of lambda seem to have apparent pupil displacement in videokeratoscope. He also noted that the best central reference point for refractive surgery should be the line of sight, even in eyes with apparent pupil displacement observed by videokeratoscopy.10 Those analyses were made by assuming a symmetrical eye with a centered pupil of circular shape. In contrast, the pupillary position and form of the human eye varies. We have seen patients who had entrance pupil displacement that was detected only by videokeratoscopes and caused decentered ablation by LASIK despite use of an eye-tracking system. This resulted in increased irregular astigmatism (unpublished data). From these experiences, we hypothesized that some eyes with asymmetric pupils or irises cannot be diagnosed with a typical ophthalmic examination but that these factors may influence the outcome of refractive surgery.
In this study, we examined the location of the pupil and tilting angle of the iris in a large number of Japanese myopic patients. We also investigated the postoperative outcome of LASIK surgery using an eye-tracking system, with respect for the change of the best spectacle-corrected visual acuity (BSCVA). The purpose of this study was to determine the anatomic variation of the position of the entrance pupil and the iris and to detect the range for safety for laser refractive surgeries.
PATIENTS AND METHODS
The study included 2280 eyes of 1144 patients (658 men and 486 women, mean age 34.3 years ± 9.8 (SD), who had a mean myopia of −5.72 ± 2.52 diopters [D]) and who had myopic LASIK surgery at the Minamiaoyama Eye Clinic. At the first visit, no eyes revealed abnormal findings such as deformation of the pupil or the iris by routine ophthalmologic examination. Preoperative BSCVA in all patients was 20/20 or better.
To study the effect of pupil decentration and tilting angle of the iris on the clinical outcome of refractive surgery, 901 eyes of 467 patients (age 34.4 ± 8.8 years, −5.56 ± 2.21 D) who had LASIK and regularly attended follow-up examinations for 6 months or more were recruited from the subjects in this study. The LASIK surgeries were performed using the LSK-One (Moria) or MK-2000 (Nidek) microkeratomes and the EC-5000 excimer laser (Nidek), which had an eye-tracking system. Mydriatic or myopic agents were not used during surgery.
Postoperative examinations were performed 1 day, 1 week, and 1, 3, and 6 months postoperatively. Eyes that repeatedly had a 1 line decrease of BSCVA or more than 2 lines on at least 1 occasion without obvious cause (ie, flap folds or corneal epithelial defects) were defined as eyes with decreased postoperative BSCVA.
Distance of Pupil Decentration
Pupillary margin and center were displayed on the apex plane anterior lens and iris modes by Orbscan (Canon). The distance between corneal apex and pupillary center on this map was measured, and the actual distance of pupil decentration was calculated.
Orbscan indicates angle kappa, which was calculated by taking the plane of the iris, locating the perpendicular line at the center of the pupil, and locating where the line intersects the cornea. Therefore, this is not the same as what is generally understood to be angle kappa. This value was used as an approximation to the tilting angle of the iris.
Multivariate Analysis of Risk Factors for Decreased Postoperative BSCVA
Risk factors for decreased postoperative BSCVA were investigated by multiple analysis of variance (ANOVA). Patient age, refractive power, corneal power, pupil decentration, and tilting angle of the iris were investigated, and P values of 0.05 were considered statistically significant.
The pupil decentration in all eyes ranged between 0 and 0.95 mm (mean 0.19 ± 0.11 mm). A detailed list of pupil decentration in this series is shown in Table 1. Pupils in 1722 eyes (75.5% of all eyes) were located superior (0 to 180 degrees) to the corneal apex, whereas those in 558 eyes (24.5%) were located inferiorly (180 to 360 degrees) (Figure 1).
Tilting Angle of the Iris
The tilting angle of the iris in all patients ranged between 0.19 and 12.69 degrees (mean 4.06 ± 1.41 degrees). A detailed list of tilting angles of the iris in this series is shown in Table 2. In 1839 eyes (80.7%), tilting of the iris was toward the temporal side; in 441 eyes (19.3%), the iris tilted nasally (Figure 2).
Correlation Between Pupil Decentration and Tilting Angle of the Iris
There was no statistical significant correlation between PD and TI (Figure 3; r = −.0256, r2 = .0007, P = .2224).
Multiple Analysis for Possible Risk Factors for Postoperative Decreased BSCVA
Thirty-four eyes of 30 patients had decreased postoperative BSCVA. Multiple ANOVA showed that refractive power, tilting angle of the iris, and pupil decentration were significant for reduction of postoperative BSCVA (Table 3). The patients' age and corneal power did not appear to be important risk factors.
In this study, we chose the corneal apex as the standard point for measurement because it is automatically measurable with both Orbscan and TMS-II (Tomey). This analysis showed that the center of the entrance pupil in many healthy eyes was dislocated approximately 0.2 mm from the corneal apex and that the iris plane was tilted approximately 4 degrees from the pupillary axis. Dislocation of the entrance pupil and tilting angle of the iris were not correlated. Neither abnormality was detected by typical ophthalmologic examination but only through topographic examination using Orbscan.
Topographic analysis showed that 96% of eyes in this study had a pupil decentration less than 0.4 mm; 80% were dislocated superiorly relative to the corneal apex. One possible explanation for this may be that the human cornea protrudes more in the inferior quadrant, probably because of gravity or pressure from the upper tarsus. Several investigators have suggested that the tension of the tarsus or upper eyelid influences the corneal shape.11–14 Those mechanical forces may displace the corneal apex slightly inferiorly, resulting in relative superior location of the pupil. Similar analysis of tilting angle of the iris showed that 93% of eyes had tilting angle of the iris less than 6 degrees and the majority of eyes examined had a shift of tilting angle of the iris temporally. The eyes with tilting angle of the iris above 6 degrees were observed in 7.3% of eyes. We believe that eyes with pupil decentration greater than 0.4 mm or tilting angle of the iris greater than 6 degrees could be categorized as eyes with extra large pupil decentration or eyes with extra large tilting angle of the iris.
Both pupil decentration and tilting angle of the iris significantly affect BSCVA and refractive power after LASIK. We think that present theories for centering the ablation may not be applicable in eyes with greater pupil decentration, especially greater than 0.4 or 0.5 mm. Some of these eyes may have a line of sight that is largely different from the real visual axis, possibly resulting in subclinical decentration of laser ablation. It may be better to recommend that those patients postpone refractive surgery, especially in the presence of high myopia.
In eyes with greater tilting angle of the iris, precise positioning of the globe during refractive surgeries may be difficult. For example, the EC-5000 excimer laser is designed to equip cross-illuminations that are 2 parallel slit lights. This equipment was basically developed to focus the laser spots on the ablated cornea, and surgeons can also estimate the eye position by determining whether the 2 light lines are parallel on the iris. In cases with a large tilting angle of the iris, the light lines on the iris do not seem to be parallel to each other. If the surgeon adjusted the eye position to make the 2 lines parallel, the globe could rotate, resulting in eccentric ablation on the cornea. It is beneficial for surgeons to know the patient's tilting angle of the iris before surgery.
In conclusion, we have demonstrated that typical ophthalmologic examination may not detect pupils decentered from the corneal apex by more than 0.4 mm or irises tilted by more than 6 degrees. Topographic examination such as that using the Orbscan device can be used in such cases. It is important to detect pupil decentration and tilting angle of the iris prior to LASIK, especially in patients with high myopia, because such cases may not be suitable for surgery without first determining the true visual axis. Further investigation in this field that considers factors such as pupillary dislocation with pupil size changes and influences on postoperative changes in higher-order aberration is essential.
1. Krueger RR. In perspective: eye tracking and autonomous laser radar. J Refract Surg 1999; 15:145-149
2. Gobbi PG, Carones F, Brancato R, et al. Automatic eye tracker for excimer laser photorefractive keratectomy. J Refract Surg 1995; 11(3 suppl):S337-S342
3. Tsai YY, Lin JM. Ablation centration after active eye-tracker-assisted photorefractive keratectomy and laser in situ keratomileusis. J Cataract Refract Surg 2000; 26:28-34
4. Taylor NM, Eikelboom RH, van Sarloos PP, Reid PG. Determining the accuracy of an eye tracking system for laser refractive surgery. J Refract Surg 2000; 16:S643-S646
5. Mrochen M, Eldine MS, Kaemmerer M, Seiler T, Hüts W. Improvement in photorefractive corneal laser surgery results using an active eye-tracking system. J Cataract Refract Surg 2001; 27:1000-1006
6. Uozato H, Guyton DL. Centering corneal surgical procedures. Am J Ophthalmol 1987; 103:264-275
7. Enoch JM, Laties AM. An analysis of retinal receptor orientation. Invest Ophthalmol 1971; 10:959-970
8. Bonds AB, MacLeod DIA. A displaced Stiles-Crawford effect associated with an eccentric pupil. Invest Ophthalmol Vis Sci 1978; 17:754-761
9. Mandell RB. Apparent pupil displacement in videokeratography. CLAO J 1994; 20:123-127
10. Mandell RB. Locating the corneal sighting center from videokeratography. J Refract Surg 1995; 11:253-259
11. Donnenfeld ED, Perry HD, Gibralter RP, Ingraham HJ, Udell IJ. Keratoconus associated with floppy eyelid syndrome. Ophthalmology 1991; 98:1674-1678
12. Negris R. Floppy eyelid syndrome associated with keratoconus. J Am Optom Assoc 1992; 63:316-319
13. Kim T, Khosla-Gupta B, Debacker C. Blepharoptosis-induced superior keratoconus. Am J Ophthalmol 2000; 130:232-234
14. Lieberman DM, Grierson JW. The lids influence on corneal shape. Cornea 2000; 19:336-342