- Email: [email protected]
A Prospective Evaluation of Corneal Curvature Changes with 3.0- to 3.5-mm Corneal Tunnel Phacoemulsification
A Prospective Evaluation of Corneal Curvature Changes with 3.0.. to 3.5..mm Corneal Tunnel Phacoemulsification Daniel A. Long, MD, l Monica L. Monica, MD, PhD2 Purpose: To document the effects and results of cataract surgery through a corneal tunnel incision centered on the steepest corneal meridian. Methods: The authors prospectively evaluated their first 172 eyes that received a 3.5-, 3.2-, or 3.0-mm corneal tunnel incision placed on the steepest corneal meridian followed by a plate haptic silicone lens implant. All surgeries were done by the same surgeon, and all eyes have completed a 12-month postoperative period. The keratometric changes were analyzed with the HOlladay-Cravy-Koch method. Results: The inciSion in the vertical meridian produced more meridional flattening and more astigmatic change (0.90 diopter) than the incision in the horizontal meridian (0.60 diopter). Both incisions had good clinical results. Only 3.6% required a suture. Although the 3.0-mm incisions induced less corneal change than the 3.5-mm incisions, the difference was not usually statistically significant. Conclusion: Three-plane corneal tunnel incisions placed on the steeper meridian are safe, consistently self-sealing, and rapidly stable; and they produce less than a 1.00diopter astigmatic change. Vertical incisions produce slightly more astigmatic change and different effects compared with horizontal incisions. Ophthalmology 1996;103:226-232
The development of foldable silicone lens implants has made it possible to insert a lens implant after phacoemulsification without enlarging the surgical wound. This has led to the development of the corneal tunnel cataract surgery wound as an easier, faster approach to cataract extraction. Because the clear cornea incision appeared to affect fewer ocular tissues, we began using this procedure in 1993. Much has been discussed, but little has been found in the literature, about corneal tunnel incisions, results, and Originally received: February 27, 1995. Revision accepted: September 26, 1995. I Private Practice, Gretna, Lousiana. 2 Clinical Associate, Tulane University Medical Center, New Orleans. Each author states that sjhe has no proprietary interest in the development or marketing of any devices, lens implants, medications, or any other product used or reported here. Correspondences to Daniel A. Long, MD, 608 Terry Pkwy, Gretna, LA 70056.
226
difficulties. Therefore, we prospectively evaluated our first 172 consecutive corneal tunnel phacoemulsification procedures. We analyzed the effect on corneal curvature from the corneal tunnel incision to see whether the incision was astigmatically neutral and to determine whether the effects were the same for all meridians. All surgical procedures were done by the same surgeon, and all eyes have completed a 12-month postoperative observation.
Materials and Methods Study subjects consisted of the first 172 consecutive procedures of corneal tunnel phacoemulsification performed on 135 patients. All implants were plate haptic silicone lenses (Chiron Vision, Model C-lO, Claremont, CA), and all were inserted with an injector cartridge system (Chiron Vision). The first patients required a 3.5-mm incision for the cartridge; as the cartridge was improved, the incision was decreased to 3.2 mm. The final largest group of pa-
Long and Monica . Corneal Curvature Changes with Corneal Tunnel Phacoemulsmcation
Table 1. The Distribution of Eyes into Groups by Incision Site and Size Group
3.5mm
3.2mm
3.0mm
Total
Horizontal Vertical Oblique
30 5 2
9 5 0
74 29 7
113 39 9
Total
37
14
110
161
tients was implanted with a cartridge that fit into an unenlarged 3.0-mm phacoemulsification wound. Incisions were placed on the steepest corneal meridian, which was determined by keratometry. To compare the results of incisions in different locations, we divided the groups arbitrarily into a vertical-incision group, with the incision site at 90 0 ± 20 0 ; a horizontal-incision group, with an incision site at 180 0 ± 20 0 ; and an oblique-incision group, with incisions from 21 0 to 69 0 and 111 0 to 159 0 • Postoperative examinations, including keratometry and refraction, were performed on day 1, week 1, and months 1, 3,6, and 12. Keratometric powers in diopters (D) and the corresponding axes were entered into LENSSTAT (Intraocular Lens and Postoperative Analysis System; Refined Logic Corporation, Dearborn, MI) and were run on an IBM-compatible computer (Packard Bell Electronics, Inc, Magna, UT). The keratometric spherical equivalent change, the keratometric astigmatic change, and with-the-wound (WTW) and against-the-wound (ATW) keratometric changes related to the incision were analyzed based on the Holladay-Cravy-Koch method. I Standard analysis of variance was done; a significant difference was indicated by P < 0.05. Data collection on an individual patient was discontinued if a secondary incisional procedure was performed.
Surgical Procedure The incision was marked for 3.5, 3.2, or 3.0 mm, depending on the size necessary for the injector cartridge. The incision was made just anterior to the insertion of the conjunctiva. A half-depth perpendicular groove incision then was made with a metal blade followed by a lamellar dissection for 3 mm into clear cornea. A stab incision was performed to the left side of the lamellar dissection, and the chamber was filled with viscoelastic material. A 3.0-mm metal keratome blade then was lubricated and inserted into the lamellar dissection. The three-plane incision was completed by pointing the tip of the keratome toward the lens and slowly inserting the blade for its full extent to produce a square 3.0-mm tunnel. After a continuous curvilinear capsulorhexis and nucleus hydrodissection, phacoemulsification and cortical aspiration were performed. Viscoelastic material then was injected to expand the capsular bag. If necessary, enlargement of the wound was done with the keratome blade to the limit of the corneal groove incision. The C-I0 lens
was inserted with the injector cartridge system. Acetylcholine was injected, and the viscoelastic agent was removed with the irrigation aspiration handpiece. Balanced salt solution was injected through the paracentesis tract to deepen the anterior chamber. The depth of the chamber was evaluated by sight, and assessment of the ability of the wound to maintain the chamber was made. Ocular pressure was checked by compression with a cellulose spear, and the wound also was tested by compression with a cellulose spear. A single, radial, 10-0 nylon suture was passed across the wound if the chamber was less than normal depth, if the globe was extremely soft to compression, o~ if there was obvious leakage from the wound; otherwIse, no suture was placed. A collagen shield soaked in antibiotic and corticosteroid was placed on the cornea. Eyes then were patched until seen in the office the following day.
Results We included the first 172 consecutive cataract surgeries intended for this technique. When the series began, every cataract surgery performed by the surgeon was to be done with the reported technique, with no exceptions or exclusions. Five eyes (2.9%) were excluded from the astigmatism analysis because of intraoperative problems: two procedures were converted to large-wound extracapsular extraction and three included lens implant exchanges at the time of cataract surgery. Six of 167 successful eyes that received a single radial suture (3.6%) also were excluded. Therefore, 161 eyes were included in the astigmatism analysis (Table 1). There is incomplete follow-up for 27 eyes (Table 2). The 3.0-mm groups had slightly less corneal change and faster stabilization than the 3.5- and 3.2- mm groups, but the differences were not statistically significant for this small group of patients. Therefore, all three incision lengths were pooled into a single horizontal-incision group, a single vertical-incision group, and a single oblique-group for comparison. Composition of the groups is shown in Table 3. The number of eyes in the obliqueincision group was too small for statistical analysis; there-
Table 2. Incomplete Follow-up Reason
No. of Patients
eVA
=
9 4
8
Death eVA Illness IOL exchange Moved Unknown Total
No. of Eyes
4
5
6 1 1 6
25
27
6 1 1
cerebrovascular accident; IOL
=
intraocular lens.
227
Ophthalmology
Volume 103, Number 2, February 1996 Table 3. Demographics
Age (yrs)
Astigmatism
Group
No.
Mean ± SD
Horizontal Vertical Oblique
113 39 9
73.5 ± 9.6 71.2 ± 7.6 71.0 ± 7.7
67 74 67
92 97 100
Total
161
72.8
69
94
Sex (%F)
Race (%W)
Eye (%OS) 52 51 78 53
Mean ± SD 0.81 ± 0.76 1.15 ± 0.78 1.03
SD = standard deviation; W = white; OS = left eye.
fore, they will not be considered further. The verticalincision group had more preoperative astigmatism than the horizontal-incision group. Otherwise, there was no statistically significant difference between the groups. Average keratometric astigmatism of the horizontal group was statistically unchanged from the average astigmatism before surgery. The vertical group showed a significant decrease in the mean astigmatism because of the incision (Table 4). Change in spherical equivalent induced by the corneal incision was near zero and clinically insignificant over the 12-month period (Tables 5 and 6). There was significantly more flattening in the vertical group compared with the horizontal group on day 1 and week 1. Flattening of the cornea is indicated by the minus sign and steepening by positive values. The magnitude of astigmatism change describes the power of a cylinder lens that would result in the postoperative corneal astigmatism ifit were held in front of the preoperative eye (Fig 1). The horizontal-incision group had significantly less induced astigmatism than the vertical-incision group on day 1 and on months 1,6, and 12 (Tables 5 and 6, Fig 2). The WTW component of the astigmatic change indicates flattening of the meridian of the incision, in general, except for the later stages of the horizontal-incision group (Tables 5 and 6, Fig 3). Analysis of variance found that the difference between the horizontal- and vertical-incision groups was statistically highly significant at all examination points. The ATW component showed a steepening in both groups at all time points, but there was no significant dif-
Table
ference between the groups (Tables 5 and 6). Also, the average axis of the induced astigmatic change is shown in Tables 5 and 6. Although the magnitude of induced astigmatism decreased slightly over the first 3 months, there was no change in the axis over the 12-month period for either group. The axis was vertical for the vertical-incision group and oblique for the horizontal-incision group.
Discussion To minimize the exacerbation of corneal astigmatism by the incision, we placed the incision on the steepest keratometric meridian, as instructed by experienced clear corneal tunnel surgeons (Fichman RA, Deutscher EM, Kershner RH; personal communication). As expected, this meridian most frequently was approximately 180 0 • Initially, operating from the side of the head was difficult and uncomfortable, but it rapidly became the preferred location. This approach avoids the brow and the eyelids and minimizes the effect of Bell phenomenon. Corneal striae and stromal fluid imbibition are more troublesome with the vertical incision, which ends closer to the center of the cornea. In addition, our results show that the temporal incision induces less astigmatic change and may attain stability more rapidly than a superior incision. Similarly, Cravy2 has reported that temporal 9.0-mm extracapsular cataract extraction incisions and temporal 6.5mm scleral tunnel incisions were more stable than superior ones. A corneal tunnel incision can be constructed in several ways, but all incisions in this study were three-planed and
4. Mean Astigmatism for the Combined Groups
Preoperative
Month 1
Month 6
Month 12
Group
Mean ± SD
No.
Mean ± SD
No.
Mean ± SD
No.
Mean ± SD
No.
Horizontal Vertical"
0.81 ± 0.76t 1.15 ± 0.78t
113 39
0.88 ± 0.77 0.76 ± 0.68t
102 37
0.84 ± 0.80 0.81 ± 0.73t
88 38
0.84 ± 0.69 0.66 ± 0.52t
90 36
SD
=
standard deviation.
* Eyes in this group had more preoperative astigmatism than the horizontal group (P < 0.05).
t
Eyes in this group had a significant decrease in mean astigmatism from preoperative levels at all postoperative periods (P < 0.05).
228
N N \Q
103 103 103 103 103
0.51* 0.56 0.58* 0.69 36
± ± ± ± ±
0.04 0.80 -0.22 0.30 149
No.
Mean ± SD
Week 1
= standard deviatio n .
35 35 35 35 35
± 0.44* ± 0.92 ± 0.69· ± 0.51 ± 13
-0.23 1.22 -0.73 0.28 90
No.
Mean ± SD
Week 1
± 0.58 ± 0.49* ± 0.68* ± 0.62 ± 31
Month 1
± ± ± ±
0.63* 0.58* 0.48 08
± 0.37
Mean ± SD -0.Q2 1.14 -0.50 0.46 90
• The re was a significant difference compared with eyes in the horizontal group (P < 0.05).
SD
35 35 35 35 35
0.44* 0.56* 0.66* 0.56 13
± ± ± ± ±
Spherical equivalent Astigmatism (magnitude) With-the-wound Against-the-wound Axis
-0.18 1.17 - 0.58 0.22 87
No.
Mean ± SD
Change
Day 1
0.12 0.73 -0.09 0.33 153
Mean ± SD
Month 1
102 102 102 102 102
No. 0.11 0.67 -0.04 0.27 155
± ± ± ± ± 0.51 0.49 0.48* 0.72 31
Mean ± SD
Month 3
86 86 86 86 86
No.
37 37 37 37 37
No.
0.01 0.88 -0.37 0.39 85
± 0.27 ± 0.66 ± 0.47* ± 0.43 ± 14
Mean ± SD
Month 3
37 37 37 37 37
No.
Table 6. Calculated Keratometric Changes for the Vertical Group
• There was a significant difference compared with eyes in the vertical group (P < 0.05).
SD = standard deviation .
98 98 98 98 98
0.54* 0.62* 0.64* 0.74 32
± ± ± ± ±
Spherical equivalent Astigmatism (magnitude) With-the-wound Against-the-wound Axis
0.04 0.96 -0.23 0.30 153
No.
Mean ± SD
Change
Day 1
Table 5. Calculated Keratometric Changes for the Horizontal Group
Month 6
± 0.41 ± 0.46* ± 0.49* ± 0.52 ± 33
0.04 0.93 -0.35 0.43 91
± 0.27 ± 0.54* ± 0.46* ± 0.33 ± 13
Mean ± SD
0.16 0.56 0.12 0.20 152
Mean ± SD
Month 6
38 38 38 38 38
No.
88 88 88 88 88
No.
± ± ± ± ±
0.41 0.47* 0.46* 0.57 31
-0.05 0.90 -0.44 0.33 90
± 0.38 ± 0.49* ± 0.51· ± 0.47 ± 11
Mean ± SD
Month 12
0.11 0.62 -0.01 0.23 154
Mean ± SD
Month 12
36 36 36 36 36
No.
90 90 90 90 90
No.
Ophthalmology
Volume 103, Number 2, February 1996 1.4 , - - - - - - - - - - - - - - - - - - - - - , 1.2 - -~-=-""----
-
+
0.8 0.6·
fInHIp
AstIgmatism
+
Aadgmll1lc Change
-
0.4
~
__
Week 1
Poat-Gpenrtlve AstIgmltlsm
Figure 1. Illustration of calculated astigmatic change.
square, with a lamellar dissection at approximately half stromal depth. The wound construction is based on principles published for scleral tunnel incisions,3-5 with the goals of astigmatism neutrality and consistent sealing without sutures. This wound can be created rapidly, easily, and consistently; stromal hydration to seal the wound was not done, and we had only 6 eyes (3.6%) of 167 that required a suture. The need for a conjunctival flap, cautery, and sutures is eliminated, and the clear cornea incision is therefore faster and less costly than a scleral tunnel wound. The Chiron Vision injection system was designed originally for 3.5-mm incisions, but it now has a 2.8- and a 2.6-mm diameter cartridge, both of which will fit into a 3.0-mm incision. A review of the literature on small-incision cataract surgery suggests that forceps delivery systems cannot be inserted through unenlarged 3.0-mm incisions. 6- 10 Pfleger et alII inserted foldable lenses through a 3.5-mm incision but did not state what insertion instrument was used. Oshika et al 12 inserted high refractive index, thin-profile, silicone lenses with forceps through a 3.2-mm incision. The ability to use an unenlarged phacoemulsification incision easily with the cartridge system seems to be an improvement over forceps delivery systems, because clear corneal incisions should be as small as possible. This study did not show a definite advantage for the 3.0-mm incision. Although the astigmatic changes were smaller for the 3.0-mm incisions, the difference was not statistically significant. A larger number of patients in each group may have produced a statistically significant difference. However, the trend favoring the smaller incision suggests that lens insertion without wound enlargement is worth the effort, and the smaller cartridge delivery system makes this possible. The horizontal-incision group had no change in average astigmatism compared with preoperative levels, but the vertical-incision group had a significant decrease in astigmatism due to a consistent against-the-rule shift. Because of preoperative astigmatism over 1.00 D, the induced astigmatic change was beneficial, leaving the vertical-incision group with an acceptable amount of postoperative astigmatism. When evaluating the corneal changes, it is important to remember that a 1.00-D change in astigmatism rep-
230
L -_ _ _
Day 1
~
___________
Month1
---
Month 3
--
Month 6
~
Month 12
Horizontal Vertical
Figure 2. Graph of the magnitude of the surgically induced astigmatic change.
resents a flattening of 0.50 D in one meridian and a steepening of 0.50 D in the meridian 90° away, if we assume a theoretical coupling ratio of -1.0. In addition, keratometry has an accuracy of ±0.25 D and ±3 °Y Therefore, both groups showed a change that is clinically small. The vertical corneal tunnel incision showed a stronger tendency to flatten the vertical meridian, coupled with ATW steepening, with the ratio of flattening to steepening 1: 1 after stabilizing. On the other hand, there appeared to be a significantly weaker tendency to flatten the horizontal meridian with a horizontal incision, coupled with a similar tendency for the ATW meridian to steepen. In fact, the horizontal-incision group steepened on both axes and induced 0.13 D of myopia. Therefore, the difference in WTW flattening explains the difference in induced astigmatism between the horizontal- and vertical-incision groups. Astigmatic change and WTW change for both groups did not stabilize until month 3 but then were stable from months 3 to 12. The drift was 0.20 to 0.30 D for both parameters. The average axis of the astigmatic change is approximately 90° for the vertical-incision group, which is an against-the-rule shift in astigmatism. Therefore, the spherocylinder lens that represents the corneal change after stabilization (average of months 3, 6, and 12) is +0.45 - 0.90 X 180, or +0.45 - 0.90 @ 90° in power notation.
0.2 , - - - - - - - - - - - - - - - - - - - - - - ,
(0.2) .. - '-=-...:=....c~___ (0.4) (0.6)
(o.8b ay 1
Week 1
Month 1
---
--
Month 3
Month 6
Month 12
Horizontal Vertical
Figure 3. Graph of the surgically induced with-the-wound change.
Long and Monica' Corneal Curvature Changes with Corneal Tunnel Phacoemulsification The spherical equivalent is zero, and the negative cylinder illustrates the net flattening of the vertical meridian. Conversion to crossed cylinder form finds (-0.45 @ 90°)(+0.45 @ 180°), which is similar to the averaged WTW (-0.38 D) and ATW (+0.38 D) data for these 3 months. The implied coupling ratio (CR) ofCR = ATW / WTW from averaged data is +0.38/-0.38, or -I, which is the theoretical coupling ratio of the cornea. For the horizontal-incision group, the axis of astigmatic change is oblique, at approximately 150°, and the spherical equivalent is increased by approximately +0.15 D. Therefore, the lens that approximates the averaged (months 3, 6, and 12) astigmatic change is -0.15 + 0.60 X 150, or -0.15 + 0.60 @ 60° in power notation. Because of the oblique axis, 25% of the astigmatism power change is applied to the 180° meridian (the axis of the incision) and 75% to the 90° meridian 14: +0.60 @ 60° = (+0.15 @ 180°)(+0.45 @ 90°). This crossed cylinder form reduces to (+0.15 D sphere)(O.OO @ 180°)(+0.30 @ 90°), which approximates the averaged data for spherical equivalent (+0.13 D), WTW (+0.02 D), and ATW (+0.23 D) changes. This is a slight myopic shift in addition to a withthe-rule shift. The implied CR from the averaged WTW and ATW data is not clinically meaningful, except to reiterate that the cornea responded differently to the horizontal incision than to the vertical incision. Koch 3 published a discussion of wound construction and argued that an astigmatically neutral wound must be made within an imaginary incisional funnel. The closer the incision is to the limbus, the shorter the incision must be. Samuelson and co-workers4 determined on cadaver eyes that a 3.0-mm scleral tunnel incision combined with a scleral recession induced approximately 0.25 D of astigmatic change but was not statistically different from a 2.0- or 2.5-mm wound. A 3.5-mm incision induced a 0.47D change and a 4.0-mm incision induced a O. 74-D change. Our results found more astigmatic change, and it is possible that a corneal tunnel incision must be even smaller than 3.0 mm to be truly neutral, especially with a vertical incision. No study has been found in the literature regarding corneal tunnel incisions or the use of the Holladay-CravyKoch method of corneal change analysis. However, in other studies, astigmatic change with the vector analysis method has been analyzed as described by Jaffe and Clayman,15 and these results are comparable to the total astigmatic change results of the Holladay-Cravy-Koch method. In Table 7, the astigmatic change is compared, as reported in several of these studies, with the results of our two groups. Our corneal tunnel incisions compare favorably with the scleral tunnel incisions. The vertical corneal tunnel incision produced slightly more astigmatic change than the superior scleral tunnel incisions, but the difference is probably clinically insignificant. The horizontal corneal tunnel incision is comparable to the unsutured superior scleral tunnel incision and better than the sutured ones in the early stages. Sutures appear to induce more early astigmatic change that decays over the first 3 months.
o
a::;
U")
rOU")
o
+1 ~ U") ..... r-
+1
U") r-
NO \0 0\
o
00
\0
o
~ ~~~~~
..c + 1 0 0 0 0 0 i:l;:: +1+1+1+1+1 ~ S ~~~~~ ~
00000
O\<"')'f"'f"Na::;r'f"\O<"')U")U")\O'f"
0000000 +1 +1 +1 +1 +1 +1 +1
~
<"')'f"O\N<"')a::;r- ..... r- ..... 'f" r- r- 0\ r-
0""'<00000
o ....
~
o
C/)
+1 ;::
S
~
\0 \0 U")
N
00 +1 +1 ~ ~ 0""'<
~
~
o
+1
<"')
a::;
o
N a::; <"')0\
""'<0 0 +1 +1 "!
'f"<"') .....
"!"!
..........
~
~
00
00000<11<.111)
... ZZZZZ>->->-
(/)
o
~~NNu")OOS
66rrirrirri~~6
rrirri
0\
II
o
if)
231
Ophthalmology
Volume 103, Number 2, February 1996
Another interesting comparison can be made to the WTW flattening in the peripheral cornea measured by photo keratoscopy in the series of Oshika et al. 12 Fifty-six percent showed no peripheral flattening, and only 6% showed flattening that extended within the central 3.0mm diameter area, which is the zone measured by standard keratometry. Their data suggest that a vertically located scleral tunnel incision produces less WTW flattening than the vertical corneal tunnel incision. Hayashi et al 16 also performed photokeratoscopy but found mild peripheral WTW flattening from the incision that resolved by month 'l in their 3.2-mm incision group. The 4.0- and 5.0-mm incision groups had not only peripheral flattening but central steepening that persisted.
Summary Our results using the described technique were favorable, even though they included our first patients, and early wounds were enlarged to 3.5 mm. The three-plane square tunnel incision was consistently self-sealing, and no adverse corneal effects were seen postoperatively. Surgery with the horizontal incision was technically easier than with the vertical incision. Induced astigmatic change was clinically small and acceptable. Placement of the incision on the steeper corneal meridian can be beneficial, if preoperative astigmatism is 0.75 D or greater. The horizontal incision produced less astigmatic change and WTW flattening than the vertical incision, but the difference was only approximately 0.25 D. Also, the 3.0-mm wounds produced slightly less astigmatic change and WTW flattening than the 3.5- and 3.2-mm incisions, but the difference was not statistically significant. Significantly larger group size may show a difference. Lens implant insertion with a cartridge system allows surgery through an unenlarged 3.0-mm incision, which is preferable with the clear cornea location.
References I. Holladay n, Cravy TV, Koch ~O . Calculating the surgically induced refractive change following ocular surgery. J Cataract Refract Surg 1992; 18:429-43.
232
2. Cravy TV. Routine use of a lateral approach to cataract extraction to achieve rapid and sustained stabilization of postoperative astigmatism. J Cataract Refract Surg 1991;17: 415-23. 3. Koch PS. Structural analysis of cataract incision construction. J Cataract Refract Surg 1991; 17 Suppl:66 1-7 . 4. Samuelson SW, Koch DO, Kugien Cc. Determination of maximal incision length for true small-incision surgery. Ophthalmic Surg 1991 ;22:204-7. 5. Ernest PH, Kiessling LA, Lavery KT. Relative strength of cataract incisions in cadaver eyes. J Cataract Refract Surg 1991;17 Suppl:668-71. 6. Brint SF, Ostrick DM, Bryan JE. Keratometric cylinder and visual performance following phacoemulsification and implantation with silicone small-incision or poly(methyl methacrylate) intraocular lenses. J Cataract Refract Surg 1991; 17:32-6. 7. Leen MM, Ho CC, Yanoff M. Association between surgically-induced astigmatism and cataract incision size in the early postoperative period. Ophthalmic Surg 1993;24:58692. 8. Masket S. One year postoperative astigmatic comparison of sutured and unsutured 4.0 mm scleral pocket incisions. J Cataract Refract Surg 1993;19:453-6. 9. Steinert RF, Brint SF, White SM, Fine IH. Astigmatism after small incision cataract surgery: A prospective, randomized, multicenter comparison of 4- and 6.5-mm incisions. Ophthalmology 1991 ;98:417 -24. 10. Utrata PJ, Sanders DR, DeLuca M, et aI. Small incision surgery with the ST AAR Elastimide three-piece posterior chamber intraocular lens. J Cataract Refract Surg 1994;20: 426-31. II. Pfleger T, Scholz U, Skorpik C. Postoperative astigmatism after no-stitch, small incision cataract surgery with 3.5 mm and 4.5 mm incisions. J Cataract Refract Surg 1994;20:4005. 12. Oshika T, Tsuboi S, Yaguchi S, et al. Comparative study of intraocular lens implantation through 3.2- and 5.5-mm incisions. Ophthalmology 1994; 10 I: 1183-90. 13. Michaels DO. Visual optics and refraction: a clinical approach. St. Louis: Mosby, 1980;346. 14. Michaels OD. Visual optics and refraction: a clinical approach. St. Louis: Mosby, 1985;52-4. IS . Jaffe NS, Clayman HM. The pathophysiology of corneal astigmatism after cataract extraction. Trans Am Acad OphthalmolOtolaryngol 1975;79:615-30. 16. Hayashi K, Hayashi H, Nakao F, Hayashi F. The correlation between incision size and corneal shape changes in sutureless cataract surgery. Ophthalmology 1995;102:550-6.
The development of foldable silicone lens implants has made it possible to insert a lens implant after phacoemulsification without enlarging the surgical wound. This has led to the development of the corneal tunnel cataract surgery wound as an easier, faster approach to cataract extraction. Because the clear cornea incision appeared to affect fewer ocular tissues, we began using this procedure in 1993. Much has been discussed, but little has been found in the literature, about corneal tunnel incisions, results, and Originally received: February 27, 1995. Revision accepted: September 26, 1995. I Private Practice, Gretna, Lousiana. 2 Clinical Associate, Tulane University Medical Center, New Orleans. Each author states that sjhe has no proprietary interest in the development or marketing of any devices, lens implants, medications, or any other product used or reported here. Correspondences to Daniel A. Long, MD, 608 Terry Pkwy, Gretna, LA 70056.
226
difficulties. Therefore, we prospectively evaluated our first 172 consecutive corneal tunnel phacoemulsification procedures. We analyzed the effect on corneal curvature from the corneal tunnel incision to see whether the incision was astigmatically neutral and to determine whether the effects were the same for all meridians. All surgical procedures were done by the same surgeon, and all eyes have completed a 12-month postoperative observation.
Materials and Methods Study subjects consisted of the first 172 consecutive procedures of corneal tunnel phacoemulsification performed on 135 patients. All implants were plate haptic silicone lenses (Chiron Vision, Model C-lO, Claremont, CA), and all were inserted with an injector cartridge system (Chiron Vision). The first patients required a 3.5-mm incision for the cartridge; as the cartridge was improved, the incision was decreased to 3.2 mm. The final largest group of pa-
Long and Monica . Corneal Curvature Changes with Corneal Tunnel Phacoemulsmcation
Table 1. The Distribution of Eyes into Groups by Incision Site and Size Group
3.5mm
3.2mm
3.0mm
Total
Horizontal Vertical Oblique
30 5 2
9 5 0
74 29 7
113 39 9
Total
37
14
110
161
tients was implanted with a cartridge that fit into an unenlarged 3.0-mm phacoemulsification wound. Incisions were placed on the steepest corneal meridian, which was determined by keratometry. To compare the results of incisions in different locations, we divided the groups arbitrarily into a vertical-incision group, with the incision site at 90 0 ± 20 0 ; a horizontal-incision group, with an incision site at 180 0 ± 20 0 ; and an oblique-incision group, with incisions from 21 0 to 69 0 and 111 0 to 159 0 • Postoperative examinations, including keratometry and refraction, were performed on day 1, week 1, and months 1, 3,6, and 12. Keratometric powers in diopters (D) and the corresponding axes were entered into LENSSTAT (Intraocular Lens and Postoperative Analysis System; Refined Logic Corporation, Dearborn, MI) and were run on an IBM-compatible computer (Packard Bell Electronics, Inc, Magna, UT). The keratometric spherical equivalent change, the keratometric astigmatic change, and with-the-wound (WTW) and against-the-wound (ATW) keratometric changes related to the incision were analyzed based on the Holladay-Cravy-Koch method. I Standard analysis of variance was done; a significant difference was indicated by P < 0.05. Data collection on an individual patient was discontinued if a secondary incisional procedure was performed.
Surgical Procedure The incision was marked for 3.5, 3.2, or 3.0 mm, depending on the size necessary for the injector cartridge. The incision was made just anterior to the insertion of the conjunctiva. A half-depth perpendicular groove incision then was made with a metal blade followed by a lamellar dissection for 3 mm into clear cornea. A stab incision was performed to the left side of the lamellar dissection, and the chamber was filled with viscoelastic material. A 3.0-mm metal keratome blade then was lubricated and inserted into the lamellar dissection. The three-plane incision was completed by pointing the tip of the keratome toward the lens and slowly inserting the blade for its full extent to produce a square 3.0-mm tunnel. After a continuous curvilinear capsulorhexis and nucleus hydrodissection, phacoemulsification and cortical aspiration were performed. Viscoelastic material then was injected to expand the capsular bag. If necessary, enlargement of the wound was done with the keratome blade to the limit of the corneal groove incision. The C-I0 lens
was inserted with the injector cartridge system. Acetylcholine was injected, and the viscoelastic agent was removed with the irrigation aspiration handpiece. Balanced salt solution was injected through the paracentesis tract to deepen the anterior chamber. The depth of the chamber was evaluated by sight, and assessment of the ability of the wound to maintain the chamber was made. Ocular pressure was checked by compression with a cellulose spear, and the wound also was tested by compression with a cellulose spear. A single, radial, 10-0 nylon suture was passed across the wound if the chamber was less than normal depth, if the globe was extremely soft to compression, o~ if there was obvious leakage from the wound; otherwIse, no suture was placed. A collagen shield soaked in antibiotic and corticosteroid was placed on the cornea. Eyes then were patched until seen in the office the following day.
Results We included the first 172 consecutive cataract surgeries intended for this technique. When the series began, every cataract surgery performed by the surgeon was to be done with the reported technique, with no exceptions or exclusions. Five eyes (2.9%) were excluded from the astigmatism analysis because of intraoperative problems: two procedures were converted to large-wound extracapsular extraction and three included lens implant exchanges at the time of cataract surgery. Six of 167 successful eyes that received a single radial suture (3.6%) also were excluded. Therefore, 161 eyes were included in the astigmatism analysis (Table 1). There is incomplete follow-up for 27 eyes (Table 2). The 3.0-mm groups had slightly less corneal change and faster stabilization than the 3.5- and 3.2- mm groups, but the differences were not statistically significant for this small group of patients. Therefore, all three incision lengths were pooled into a single horizontal-incision group, a single vertical-incision group, and a single oblique-group for comparison. Composition of the groups is shown in Table 3. The number of eyes in the obliqueincision group was too small for statistical analysis; there-
Table 2. Incomplete Follow-up Reason
No. of Patients
eVA
=
9 4
8
Death eVA Illness IOL exchange Moved Unknown Total
No. of Eyes
4
5
6 1 1 6
25
27
6 1 1
cerebrovascular accident; IOL
=
intraocular lens.
227
Ophthalmology
Volume 103, Number 2, February 1996 Table 3. Demographics
Age (yrs)
Astigmatism
Group
No.
Mean ± SD
Horizontal Vertical Oblique
113 39 9
73.5 ± 9.6 71.2 ± 7.6 71.0 ± 7.7
67 74 67
92 97 100
Total
161
72.8
69
94
Sex (%F)
Race (%W)
Eye (%OS) 52 51 78 53
Mean ± SD 0.81 ± 0.76 1.15 ± 0.78 1.03
SD = standard deviation; W = white; OS = left eye.
fore, they will not be considered further. The verticalincision group had more preoperative astigmatism than the horizontal-incision group. Otherwise, there was no statistically significant difference between the groups. Average keratometric astigmatism of the horizontal group was statistically unchanged from the average astigmatism before surgery. The vertical group showed a significant decrease in the mean astigmatism because of the incision (Table 4). Change in spherical equivalent induced by the corneal incision was near zero and clinically insignificant over the 12-month period (Tables 5 and 6). There was significantly more flattening in the vertical group compared with the horizontal group on day 1 and week 1. Flattening of the cornea is indicated by the minus sign and steepening by positive values. The magnitude of astigmatism change describes the power of a cylinder lens that would result in the postoperative corneal astigmatism ifit were held in front of the preoperative eye (Fig 1). The horizontal-incision group had significantly less induced astigmatism than the vertical-incision group on day 1 and on months 1,6, and 12 (Tables 5 and 6, Fig 2). The WTW component of the astigmatic change indicates flattening of the meridian of the incision, in general, except for the later stages of the horizontal-incision group (Tables 5 and 6, Fig 3). Analysis of variance found that the difference between the horizontal- and vertical-incision groups was statistically highly significant at all examination points. The ATW component showed a steepening in both groups at all time points, but there was no significant dif-
Table
ference between the groups (Tables 5 and 6). Also, the average axis of the induced astigmatic change is shown in Tables 5 and 6. Although the magnitude of induced astigmatism decreased slightly over the first 3 months, there was no change in the axis over the 12-month period for either group. The axis was vertical for the vertical-incision group and oblique for the horizontal-incision group.
Discussion To minimize the exacerbation of corneal astigmatism by the incision, we placed the incision on the steepest keratometric meridian, as instructed by experienced clear corneal tunnel surgeons (Fichman RA, Deutscher EM, Kershner RH; personal communication). As expected, this meridian most frequently was approximately 180 0 • Initially, operating from the side of the head was difficult and uncomfortable, but it rapidly became the preferred location. This approach avoids the brow and the eyelids and minimizes the effect of Bell phenomenon. Corneal striae and stromal fluid imbibition are more troublesome with the vertical incision, which ends closer to the center of the cornea. In addition, our results show that the temporal incision induces less astigmatic change and may attain stability more rapidly than a superior incision. Similarly, Cravy2 has reported that temporal 9.0-mm extracapsular cataract extraction incisions and temporal 6.5mm scleral tunnel incisions were more stable than superior ones. A corneal tunnel incision can be constructed in several ways, but all incisions in this study were three-planed and
4. Mean Astigmatism for the Combined Groups
Preoperative
Month 1
Month 6
Month 12
Group
Mean ± SD
No.
Mean ± SD
No.
Mean ± SD
No.
Mean ± SD
No.
Horizontal Vertical"
0.81 ± 0.76t 1.15 ± 0.78t
113 39
0.88 ± 0.77 0.76 ± 0.68t
102 37
0.84 ± 0.80 0.81 ± 0.73t
88 38
0.84 ± 0.69 0.66 ± 0.52t
90 36
SD
=
standard deviation.
* Eyes in this group had more preoperative astigmatism than the horizontal group (P < 0.05).
t
Eyes in this group had a significant decrease in mean astigmatism from preoperative levels at all postoperative periods (P < 0.05).
228
N N \Q
103 103 103 103 103
0.51* 0.56 0.58* 0.69 36
± ± ± ± ±
0.04 0.80 -0.22 0.30 149
No.
Mean ± SD
Week 1
= standard deviatio n .
35 35 35 35 35
± 0.44* ± 0.92 ± 0.69· ± 0.51 ± 13
-0.23 1.22 -0.73 0.28 90
No.
Mean ± SD
Week 1
± 0.58 ± 0.49* ± 0.68* ± 0.62 ± 31
Month 1
± ± ± ±
0.63* 0.58* 0.48 08
± 0.37
Mean ± SD -0.Q2 1.14 -0.50 0.46 90
• The re was a significant difference compared with eyes in the horizontal group (P < 0.05).
SD
35 35 35 35 35
0.44* 0.56* 0.66* 0.56 13
± ± ± ± ±
Spherical equivalent Astigmatism (magnitude) With-the-wound Against-the-wound Axis
-0.18 1.17 - 0.58 0.22 87
No.
Mean ± SD
Change
Day 1
0.12 0.73 -0.09 0.33 153
Mean ± SD
Month 1
102 102 102 102 102
No. 0.11 0.67 -0.04 0.27 155
± ± ± ± ± 0.51 0.49 0.48* 0.72 31
Mean ± SD
Month 3
86 86 86 86 86
No.
37 37 37 37 37
No.
0.01 0.88 -0.37 0.39 85
± 0.27 ± 0.66 ± 0.47* ± 0.43 ± 14
Mean ± SD
Month 3
37 37 37 37 37
No.
Table 6. Calculated Keratometric Changes for the Vertical Group
• There was a significant difference compared with eyes in the vertical group (P < 0.05).
SD = standard deviation .
98 98 98 98 98
0.54* 0.62* 0.64* 0.74 32
± ± ± ± ±
Spherical equivalent Astigmatism (magnitude) With-the-wound Against-the-wound Axis
0.04 0.96 -0.23 0.30 153
No.
Mean ± SD
Change
Day 1
Table 5. Calculated Keratometric Changes for the Horizontal Group
Month 6
± 0.41 ± 0.46* ± 0.49* ± 0.52 ± 33
0.04 0.93 -0.35 0.43 91
± 0.27 ± 0.54* ± 0.46* ± 0.33 ± 13
Mean ± SD
0.16 0.56 0.12 0.20 152
Mean ± SD
Month 6
38 38 38 38 38
No.
88 88 88 88 88
No.
± ± ± ± ±
0.41 0.47* 0.46* 0.57 31
-0.05 0.90 -0.44 0.33 90
± 0.38 ± 0.49* ± 0.51· ± 0.47 ± 11
Mean ± SD
Month 12
0.11 0.62 -0.01 0.23 154
Mean ± SD
Month 12
36 36 36 36 36
No.
90 90 90 90 90
No.
Ophthalmology
Volume 103, Number 2, February 1996 1.4 , - - - - - - - - - - - - - - - - - - - - - , 1.2 - -~-=-""----
-
+
0.8 0.6·
fInHIp
AstIgmatism
+
Aadgmll1lc Change
-
0.4
~
__
Week 1
Poat-Gpenrtlve AstIgmltlsm
Figure 1. Illustration of calculated astigmatic change.
square, with a lamellar dissection at approximately half stromal depth. The wound construction is based on principles published for scleral tunnel incisions,3-5 with the goals of astigmatism neutrality and consistent sealing without sutures. This wound can be created rapidly, easily, and consistently; stromal hydration to seal the wound was not done, and we had only 6 eyes (3.6%) of 167 that required a suture. The need for a conjunctival flap, cautery, and sutures is eliminated, and the clear cornea incision is therefore faster and less costly than a scleral tunnel wound. The Chiron Vision injection system was designed originally for 3.5-mm incisions, but it now has a 2.8- and a 2.6-mm diameter cartridge, both of which will fit into a 3.0-mm incision. A review of the literature on small-incision cataract surgery suggests that forceps delivery systems cannot be inserted through unenlarged 3.0-mm incisions. 6- 10 Pfleger et alII inserted foldable lenses through a 3.5-mm incision but did not state what insertion instrument was used. Oshika et al 12 inserted high refractive index, thin-profile, silicone lenses with forceps through a 3.2-mm incision. The ability to use an unenlarged phacoemulsification incision easily with the cartridge system seems to be an improvement over forceps delivery systems, because clear corneal incisions should be as small as possible. This study did not show a definite advantage for the 3.0-mm incision. Although the astigmatic changes were smaller for the 3.0-mm incisions, the difference was not statistically significant. A larger number of patients in each group may have produced a statistically significant difference. However, the trend favoring the smaller incision suggests that lens insertion without wound enlargement is worth the effort, and the smaller cartridge delivery system makes this possible. The horizontal-incision group had no change in average astigmatism compared with preoperative levels, but the vertical-incision group had a significant decrease in astigmatism due to a consistent against-the-rule shift. Because of preoperative astigmatism over 1.00 D, the induced astigmatic change was beneficial, leaving the vertical-incision group with an acceptable amount of postoperative astigmatism. When evaluating the corneal changes, it is important to remember that a 1.00-D change in astigmatism rep-
230
L -_ _ _
Day 1
~
___________
Month1
---
Month 3
--
Month 6
~
Month 12
Horizontal Vertical
Figure 2. Graph of the magnitude of the surgically induced astigmatic change.
resents a flattening of 0.50 D in one meridian and a steepening of 0.50 D in the meridian 90° away, if we assume a theoretical coupling ratio of -1.0. In addition, keratometry has an accuracy of ±0.25 D and ±3 °Y Therefore, both groups showed a change that is clinically small. The vertical corneal tunnel incision showed a stronger tendency to flatten the vertical meridian, coupled with ATW steepening, with the ratio of flattening to steepening 1: 1 after stabilizing. On the other hand, there appeared to be a significantly weaker tendency to flatten the horizontal meridian with a horizontal incision, coupled with a similar tendency for the ATW meridian to steepen. In fact, the horizontal-incision group steepened on both axes and induced 0.13 D of myopia. Therefore, the difference in WTW flattening explains the difference in induced astigmatism between the horizontal- and vertical-incision groups. Astigmatic change and WTW change for both groups did not stabilize until month 3 but then were stable from months 3 to 12. The drift was 0.20 to 0.30 D for both parameters. The average axis of the astigmatic change is approximately 90° for the vertical-incision group, which is an against-the-rule shift in astigmatism. Therefore, the spherocylinder lens that represents the corneal change after stabilization (average of months 3, 6, and 12) is +0.45 - 0.90 X 180, or +0.45 - 0.90 @ 90° in power notation.
0.2 , - - - - - - - - - - - - - - - - - - - - - - ,
(0.2) .. - '-=-...:=....c~___ (0.4) (0.6)
(o.8b ay 1
Week 1
Month 1
---
--
Month 3
Month 6
Month 12
Horizontal Vertical
Figure 3. Graph of the surgically induced with-the-wound change.
Long and Monica' Corneal Curvature Changes with Corneal Tunnel Phacoemulsification The spherical equivalent is zero, and the negative cylinder illustrates the net flattening of the vertical meridian. Conversion to crossed cylinder form finds (-0.45 @ 90°)(+0.45 @ 180°), which is similar to the averaged WTW (-0.38 D) and ATW (+0.38 D) data for these 3 months. The implied coupling ratio (CR) ofCR = ATW / WTW from averaged data is +0.38/-0.38, or -I, which is the theoretical coupling ratio of the cornea. For the horizontal-incision group, the axis of astigmatic change is oblique, at approximately 150°, and the spherical equivalent is increased by approximately +0.15 D. Therefore, the lens that approximates the averaged (months 3, 6, and 12) astigmatic change is -0.15 + 0.60 X 150, or -0.15 + 0.60 @ 60° in power notation. Because of the oblique axis, 25% of the astigmatism power change is applied to the 180° meridian (the axis of the incision) and 75% to the 90° meridian 14: +0.60 @ 60° = (+0.15 @ 180°)(+0.45 @ 90°). This crossed cylinder form reduces to (+0.15 D sphere)(O.OO @ 180°)(+0.30 @ 90°), which approximates the averaged data for spherical equivalent (+0.13 D), WTW (+0.02 D), and ATW (+0.23 D) changes. This is a slight myopic shift in addition to a withthe-rule shift. The implied CR from the averaged WTW and ATW data is not clinically meaningful, except to reiterate that the cornea responded differently to the horizontal incision than to the vertical incision. Koch 3 published a discussion of wound construction and argued that an astigmatically neutral wound must be made within an imaginary incisional funnel. The closer the incision is to the limbus, the shorter the incision must be. Samuelson and co-workers4 determined on cadaver eyes that a 3.0-mm scleral tunnel incision combined with a scleral recession induced approximately 0.25 D of astigmatic change but was not statistically different from a 2.0- or 2.5-mm wound. A 3.5-mm incision induced a 0.47D change and a 4.0-mm incision induced a O. 74-D change. Our results found more astigmatic change, and it is possible that a corneal tunnel incision must be even smaller than 3.0 mm to be truly neutral, especially with a vertical incision. No study has been found in the literature regarding corneal tunnel incisions or the use of the Holladay-CravyKoch method of corneal change analysis. However, in other studies, astigmatic change with the vector analysis method has been analyzed as described by Jaffe and Clayman,15 and these results are comparable to the total astigmatic change results of the Holladay-Cravy-Koch method. In Table 7, the astigmatic change is compared, as reported in several of these studies, with the results of our two groups. Our corneal tunnel incisions compare favorably with the scleral tunnel incisions. The vertical corneal tunnel incision produced slightly more astigmatic change than the superior scleral tunnel incisions, but the difference is probably clinically insignificant. The horizontal corneal tunnel incision is comparable to the unsutured superior scleral tunnel incision and better than the sutured ones in the early stages. Sutures appear to induce more early astigmatic change that decays over the first 3 months.
o
a::;
U")
rOU")
o
+1 ~ U") ..... r-
+1
U") r-
NO \0 0\
o
00
\0
o
~ ~~~~~
..c + 1 0 0 0 0 0 i:l;:: +1+1+1+1+1 ~ S ~~~~~ ~
00000
O\<"')'f"'f"Na::;r'f"\O<"')U")U")\O'f"
0000000 +1 +1 +1 +1 +1 +1 +1
~
<"')'f"O\N<"')a::;r- ..... r- ..... 'f" r- r- 0\ r-
0""'<00000
o ....
~
o
C/)
+1 ;::
S
~
\0 \0 U")
N
00 +1 +1 ~ ~ 0""'<
~
~
o
+1
<"')
a::;
o
N a::; <"')0\
""'<0 0 +1 +1 "!
'f"<"') .....
"!"!
..........
~
~
00
00000<11<.111)
... ZZZZZ>->->-
(/)
o
~~NNu")OOS
66rrirrirri~~6
rrirri
0\
II
o
if)
231
Ophthalmology
Volume 103, Number 2, February 1996
Another interesting comparison can be made to the WTW flattening in the peripheral cornea measured by photo keratoscopy in the series of Oshika et al. 12 Fifty-six percent showed no peripheral flattening, and only 6% showed flattening that extended within the central 3.0mm diameter area, which is the zone measured by standard keratometry. Their data suggest that a vertically located scleral tunnel incision produces less WTW flattening than the vertical corneal tunnel incision. Hayashi et al 16 also performed photokeratoscopy but found mild peripheral WTW flattening from the incision that resolved by month 'l in their 3.2-mm incision group. The 4.0- and 5.0-mm incision groups had not only peripheral flattening but central steepening that persisted.
Summary Our results using the described technique were favorable, even though they included our first patients, and early wounds were enlarged to 3.5 mm. The three-plane square tunnel incision was consistently self-sealing, and no adverse corneal effects were seen postoperatively. Surgery with the horizontal incision was technically easier than with the vertical incision. Induced astigmatic change was clinically small and acceptable. Placement of the incision on the steeper corneal meridian can be beneficial, if preoperative astigmatism is 0.75 D or greater. The horizontal incision produced less astigmatic change and WTW flattening than the vertical incision, but the difference was only approximately 0.25 D. Also, the 3.0-mm wounds produced slightly less astigmatic change and WTW flattening than the 3.5- and 3.2-mm incisions, but the difference was not statistically significant. Significantly larger group size may show a difference. Lens implant insertion with a cartridge system allows surgery through an unenlarged 3.0-mm incision, which is preferable with the clear cornea location.
References I. Holladay n, Cravy TV, Koch ~O . Calculating the surgically induced refractive change following ocular surgery. J Cataract Refract Surg 1992; 18:429-43.
232
2. Cravy TV. Routine use of a lateral approach to cataract extraction to achieve rapid and sustained stabilization of postoperative astigmatism. J Cataract Refract Surg 1991;17: 415-23. 3. Koch PS. Structural analysis of cataract incision construction. J Cataract Refract Surg 1991; 17 Suppl:66 1-7 . 4. Samuelson SW, Koch DO, Kugien Cc. Determination of maximal incision length for true small-incision surgery. Ophthalmic Surg 1991 ;22:204-7. 5. Ernest PH, Kiessling LA, Lavery KT. Relative strength of cataract incisions in cadaver eyes. J Cataract Refract Surg 1991;17 Suppl:668-71. 6. Brint SF, Ostrick DM, Bryan JE. Keratometric cylinder and visual performance following phacoemulsification and implantation with silicone small-incision or poly(methyl methacrylate) intraocular lenses. J Cataract Refract Surg 1991; 17:32-6. 7. Leen MM, Ho CC, Yanoff M. Association between surgically-induced astigmatism and cataract incision size in the early postoperative period. Ophthalmic Surg 1993;24:58692. 8. Masket S. One year postoperative astigmatic comparison of sutured and unsutured 4.0 mm scleral pocket incisions. J Cataract Refract Surg 1993;19:453-6. 9. Steinert RF, Brint SF, White SM, Fine IH. Astigmatism after small incision cataract surgery: A prospective, randomized, multicenter comparison of 4- and 6.5-mm incisions. Ophthalmology 1991 ;98:417 -24. 10. Utrata PJ, Sanders DR, DeLuca M, et aI. Small incision surgery with the ST AAR Elastimide three-piece posterior chamber intraocular lens. J Cataract Refract Surg 1994;20: 426-31. II. Pfleger T, Scholz U, Skorpik C. Postoperative astigmatism after no-stitch, small incision cataract surgery with 3.5 mm and 4.5 mm incisions. J Cataract Refract Surg 1994;20:4005. 12. Oshika T, Tsuboi S, Yaguchi S, et al. Comparative study of intraocular lens implantation through 3.2- and 5.5-mm incisions. Ophthalmology 1994; 10 I: 1183-90. 13. Michaels DO. Visual optics and refraction: a clinical approach. St. Louis: Mosby, 1980;346. 14. Michaels OD. Visual optics and refraction: a clinical approach. St. Louis: Mosby, 1985;52-4. IS . Jaffe NS, Clayman HM. The pathophysiology of corneal astigmatism after cataract extraction. Trans Am Acad OphthalmolOtolaryngol 1975;79:615-30. 16. Hayashi K, Hayashi H, Nakao F, Hayashi F. The correlation between incision size and corneal shape changes in sutureless cataract surgery. Ophthalmology 1995;102:550-6.