Vector Analysis of Corneal and Refractive Astigmatism Changes Following Toric Pseudophakic and Toric Phakic IOL Implantation

Abstract

Purpose.: To determine the efficacy of the astigmatism correction following toric intraocular lens (IOL) and toric phakic IOL (pIOL) implantation in eyes with no previous ocular surgery and in postkeratoplasty (PKP) eyes. In addition, changes in corneal astigmatism were determined.

Methods.: Astigmatism was analyzed in 35 eyes with an AcrySof toric IOL, 35 eyes with an Artiflex toric pIOL, 50 eyes with an Artisan toric pIOL, and 40 PKP eyes with an Artisan toric pIOL. Refractive astigmatism was analyzed by using Alpins method. Surgically induced corneal astigmatism (SICA) was determined following a superior 2.2-mm, 3.4-mm, or 5.4-mm incision. Follow-up was 12 months.

Results.: Following toric IOL implantation, the index of success was 0.14 and overall residual astigmatism, 0.37 diopter (D). Following toric pIOL implantation, the index of success was 0.32 (Artiflex) and 0.18 (Artisan), and overall residual astigmatism was approximately 0.60 D. In PKP eyes, Artisan pIOLs resulted in an index of success of 0.28 and overall residual astigmatism of 1.56 D. The SICA, following 2.2-mm, 3.4-mm, 5.4-mm (normal eyes), and 5.4-mm (PKP eyes) incisions, was −0.25 ± 0.42 D (P = 0.108), −0.31 ± 0.43 D (P < 0.001), −0.48 ± 0.55 D (P < 0.001), and −0.49 ± 1.48 D (P = 0.035), respectively.

Conclusions.: Toric IOLs and pIOLs provide an effective astigmatism correction. Incorporating the SICA into the toric IOL power calculation may further increase their effectiveness. Therefore, incorporation of 0 D, −0.30 D, or −0.50 D of SICA for a 2.2-, 3.4-, or 5.4-mm superior incision, respectively, is recommended.

Introduction

Toric pseudophakic intraocular lenses (IOLs) and toric phakic IOLs (pIOLs) may be used in cataract and refractive surgery to achieve spectacle independency for distance vision. Toric pseudophakic IOLs correct corneal astigmatism and have been shown to result in good visual and refractive outcomes. ^1–4 Two types of toric pIOLs are currently available: the anterior chamber (iris-fixated) toric pIOLs and the posterior chamber toric pIOLs. Both types of lenses have been shown to be effective in the correction of moderate to high amounts of refractive astigmatism. ^5–7 However, a vector analysis to determine the efficacy of the astigmatism correction with these toric IOLs or toric pIOLs is not commonly performed. 

Toric pIOL implantation may also be performed in patients with high levels of astigmatism due to a penetrating keratoplasty (PKP). ^8–10 Approximately 30% of PKP patients have more than 5 diopters (D) of astigmatism. ¹¹ High degrees of astigmatism may not be suitable for spectacle correction owing to anisometropia. ¹² Furthermore, owing to contact lens intolerance in PKP patients, many patients require a surgical correction of astigmatism. An advantage of toric pIOL implantation in these patients is the ability to exchange and remove the pIOL in case of a future rekeratoplasty. 

An important aspect to consider in both toric IOL and toric pIOL implantation in normal eyes is the change in corneal astigmatism induced by the incision. Most commonly used toric IOLs require a 2.2-mm incision, whereas foldable or rigid toric pIOLs require a 3.4-mm or 5.4-mm incision, respectively. In PKP eyes the biomechanical wound-healing response and the change in corneal astigmatism may differ from normal eyes. 

The purpose of this study was to evaluate the vector changes in both refractive and corneal astigmatism following toric IOL and toric pIOL implantation and to determine the efficacy of the astigmatism correction. In addition, vector changes in refractive and corneal astigmatism in PKP eyes were compared to those in eyes with no previous ocular surgery. 

Methods

Study Design and Patient Population

In this retrospective case series, patients were included who underwent toric IOL or toric pIOL implantation between January 2001 and October 2010. All surgeries were performed by two experienced surgeons (N.B. and R.N.) at the Maastricht University Medical Centre, The Netherlands. The tenets of the Declaration of Helsinki were followed and patients provided informed consent. 

Inclusion criteria for toric pseudophakic IOL implantation were regular corneal astigmatism of 1.25 D or more and cataract. Exclusion criteria were tear-film abnormalities, Fuchs' endothelial dystrophy (more than 2+ guttata), or extensive visual loss due to macular disease or glaucoma. Inclusion criteria for toric pIOL implantation in normal eyes were a subjective refractive astigmatism of 1.5 D or more, a stable refraction during the previous 2 years, and unsatisfactory correction with spectacles or contact lenses. Exclusion criteria were an anterior chamber depth (ACD) of less than 3.2 mm (measured from the epithelium to the crystalline lens), endothelial cell density of less than 2000 cells/mm², abnormal iris or pupil, glaucoma, or uveitis. Inclusion criteria for toric pIOL implantation in PKP eyes were spectacle intolerance due to anisometropia and contact lens intolerance. Exclusion criteria were an ACD less than 3.2 mm, endothelial cell density less than 500 cells/mm², glaucoma, or uveitis. 

Preoperatively, all patients underwent a complete ophthalmic evaluation including subjective refraction, Snellen best-corrected distance visual acuity (CDVA), slit lamp examination, fundoscopy, applanation tonometry, optical biometry (IOLMaster; Carl Zeiss Meditec, Jena, Germany) and manual keratometry (Javal-Schiotz; Rodenstock, Dusseldorf, Germany). All patients underwent corneal topography with the Atlas (Carl Zeiss Meditec) or EyeMap (Alcon, Fort Worth, TX) Placido disk videokeratoscope. In case of pIOL implantation, patients additionally underwent noncontact specular microscopy (Noncon Robo SP-9000; Konan Medical Incorporation, Japan) and anterior segment optical coherence tomography (Visante; Carl Zeiss Meditec). 

Intraocular Lens Types

The AcrySof toric IOL (Alcon) is a foldable IOL made of hydrophobic acrylic material and has a 6.0-mm optic diameter. It is available in cylinder powers of 1.5 D to 6.0 D (SN60T3 to SN60T9). The IOL cylinder power and alignment axis were calculated on the basis of keratometry values (K-values) obtained by optical biometry and using an online calculator (available at www.acrysoftoriccalculator.com). An expected amount of surgically induced corneal astigmatism (SICA) of 0.5 D (superior incision) was incorporated in the IOL power calculation. 

The iris-fixated Artiflex toric pIOL (Ophtec BV, Groningen, The Netherlands) is a 3-piece foldable IOL consisting of a silicone 6.0-mm optic and 2 rigid poly(methyl methacrylate) (PMMA) haptics. It is available in cylinder powers of 1.0 D to 5.0 D. The iris-fixated Artisan toric pIOL (Ophtec BV) is made from nonfoldable PMMA and has a 5.0-mm optic. Available cylinder powers range from 1.0 D to 7.5 D. For both pIOLs, the lens power was calculated by using the van der Heijde formula based on the mean K-value, ACD, refractive spherical equivalent, and refractive cylinder power (corrected for a 12.0-mm vertex distance). The axis of surgical enclavation was derived from the subjective refraction. All power calculations were performed by Ophtec. 

Surgical Techniques

The marking steps required for both toric IOL and toric pIOL implantation were identical. Preoperatively, after administration of a topical anesthetic, the patient was positioned upright to correct for cyclotorsion and asked to fixate an object at a distance. Limbal reference marks were placed at 0°, 180°, and 270° (3-, 6-, and 9-o'clock positions) with a Nuijts/Lane Toric Reference Marker with bubble level (ASICO, Westmont, IL). Intraoperatively, the limbal reference marks were used to mark the alignment axis. This was done with a Mendez degree gauge (ASICO) and a Nuijts Toric Axis Marker (ASICO). 

In the case of cataract surgery with toric IOL implantation, a standard phacoemulsification was performed through a 2.2-mm superior limbal incision. In the case of refractive surgery with toric pIOL implantation, a 3.4-mm (Artiflex) or a 5.4-mm (Artisan) superior corneoscleral incision was used. Following Artiflex pIOL implantation, the wound was sutured by using two interrupted 10-0 nylon sutures. Following Artisan pIOL implantation, 5 interrupted 10-0 nylon sutures were used to close the incision. Suture removal was performed between 1 and 3 months postoperatively. 

Postoperative Examinations

At 6 and 12 months postoperatively, all patients underwent subjective refraction, measurement of Snellen uncorrected distance visual acuity (UDVA) and CDVA and corneal topography. For each patient, preoperative and postoperative corneal topography was always performed with the same device (Atlas or Eyemap). 

Data Analysis

The accuracy of the astigmatism correction was calculated by using a vector analysis according to Alpins. ¹³ In patients implanted with a toric IOL, postoperative refractive astigmatism was compared to preoperative keratometric astigmatism (IOLMaster). In patients implanted with a toric pIOL, postoperative refractive astigmatism was compared to preoperative refractive astigmatism. Target astigmatism was zero, since emmetropia was the goal in all patients. Refractive astigmatism data were calculated to the corneal plane as follows. Subjective refraction at the spectacle plane was converted into the two principal lens powers. Both lens powers were than vertexed to the corneal plane by using equation (1) and a back vertex distance of 12 mm. The difference between both lens powers at the corneal plane represents the amount of astigmatism at the corneal plane.  

 

where F_c = lens power (D) at corneal plane, F_s = lens power (D) at spectacle plane, and d = vertex distance (mm).  

Astigmatism magnitude (diopters) and axis (degrees) values were transformed into rectangular x and y coordinates and used to calculate the following vectors. ¹³ The target induced astigmatism (TIA) vector represents the change (by magnitude and axis) the surgery was intended to induce. The surgically induced astigmatism (SIA) vector is the astigmatic change the surgery actually induced. The difference vector (DV) represents the vector that enables the achieved astigmatic outcome to achieve target astigmatism. This is an absolute measure of success and is preferably zero. The magnitude of error (ME) is the arithmetic difference between the magnitudes of the SIA and TIA. The ME is positive for overcorrection and negative for undercorrection. The angle of error (AE) is the angle between the SIA and TIA vectors. The AE is positive if the achieved correction is counterclockwise to the intended axis and negative if the achieved correction is clockwise to the intended axis. The flattening effect (FE) is the magnitude of astigmatism reduction achieved at the intended meridian of treatment (TIA meridian). The flattening index is calculated by dividing the FE by the magnitude of TIA and is preferably 1.0. The correction index was calculated as the ratio of the magnitude of SIA to the magnitude of TIA. The correction index is preferably 1.0. It is greater than 1.0 if overcorrection occurred and less than 1.0 if undercorrection occurred. The index of success was calculated by dividing the magnitude of DV by the magnitude of TIA. This is a relative measure of success and is preferably zero. The overall mean magnitude of residual astigmatism and the residual astigmatism at the meridian of treatment were calculated. ¹⁴  

The SICA was defined as the vector change in corneal astigmatism between the preoperative and postoperative period. To calculate the SICA, corneal astigmatism data, obtained by corneal topography, were transformed into Cartesian coordinates (x and y coordinates) according to the method described by Holladay et al. ¹⁵ The SICA was calculated by using a standard vector analysis and is shown in double-angle minus-cylinder power plots. The standard deviations (SDs) of the x and y coordinates are displayed as an ellipse surrounding the centroid (the mean SICA) in the double-angle plots. ¹⁵ After the calculations were finished, the Cartesian coordinates were transformed back to the standard notation for astigmatism (cylinder and meridian). 

Statistics

All data were collected in an Excel database (Microsoft Office 2003; Microsoft, Redmond, WA) and analyzed with SPSS for Windows (version 16.0, SPSS Inc., Chicago, IL). Snellen UDVA and CDVA were converted into LogMAR for the mathematical and statistical calculations. Postoperative changes in astigmatism vectors within each group were analyzed using a repeated measures analysis of variance (ANOVA). A Hotelling Trace multivariate analysis of variance (MANOVA) was used to determine if vectors were significantly different from zero. Paired samples t-tests were used to analyze visual acuity and astigmatism parameters within a group. To determine if certain visual acuity or astigmatism parameters were a significant predictor for the postoperative UDVA, a multiple regression analysis was conducted. The following parameters were entered into the regression model: preoperative CDVA, ME, absolute AE, flattening index, correction index, index of success, overall residual astigmatism, and residual astigmatism at TIA meridian. A P value of less than 0.05 was considered statistically significant. 

Results

Refractive and corneal astigmatism were analyzed in 35 normal eyes with an AcrySof toric IOL, 35 normal eyes with an Artiflex toric pIOL, 50 normal eyes with an Artisan toric pIOL, and 40 PKP eyes with an Artisan toric pIOL. Patient characteristics are shown in Table 1. 

Table 1.

 

Preoperative Patient Characteristics

	Toric IOL Group
	AcrySof IOL	Artiflex pIOL	Artisan pIOL	Artisan pIOL
Population	Normal eyes	Normal eyes	Normal eyes	PKP eyes
No. of eyes	35	35	50	40
No. of patients	20	20	33	39
Age, mean ± SD	65 ± 12	45 ± 9	40 ± 10	59 ± 19
Female, %	74	74	52	56

Visual Outcomes

The preoperative and postoperative UDVA and CDVA are shown in Table 2. In all subgroups, the postoperative CDVA was significantly better than the UDVA at 6 months (P < 0.001) and 12 months (P = 0.001) follow-up. In none of the subgroups did the UDVA or CDVA change from 6 to 12 months postoperatively (P > 0.05). 

Table 2.

 

Preoperative and Postoperative LogMAR Visual Acuity

		Toric IOL Group
		AcrySof IOL	Artiflex pIOL	Artisan pIOL	Artisan pIOL
Population		Normal eyes	Normal eyes	Normal eyes	PKP eyes
Preoperatively	CDVA	0.41 ± 0.31	0.04 ± 0.12	0.12 ± 0.15	0.23 ± 0.17
6 Months postoperatively	UDVA	0.10 ± 0.10*	0.03 ± 0.11*	0.15 ± 0.18*	0.45 ± 0.24*
	CDVA	0.01 ± 0.11	−0.03 ± 0.12	0.04 ± 0.11	0.16 ± 0.14
12 Months postoperatively	UDVA	0.12 ± 0.11*	0.05 ± 0.16*	0.13 ± 0.13*	0.46 ± 0.24*
	CDVA	0.03 ± 0.11	−0.04 ± 0.10	0.04 ± 0.10	0.17 ± 0.14

*  Significantly different from CDVA at the same time point (paired samples t-test; P < 0.05).

Vector Analysis of Refractive Astigmatism

Table 3 shows the results of Alpins vector analysis based on postoperative refractive astigmatism, and Figure 1 demonstrates the variability in postoperative refractive astigmatism within each group. 

Table 3.

 

Refractive Astigmatism Outcomes at 6 and 12 Months Postoperatively in Patients Who Received an AcrySof Toric IOL, Artiflex pIOL, or Artisan pIOL

	Toric IOL Group
	AcrySof IOL	Artiflex pIOL	Artisan pIOL	Artisan pIOL
Population	Normal eyes	Normal eyes	Normal eyes	PKP eyes
Preoperatively
Preoperative astigmatism
Arithmetic mean, D	2.82	1.98	3.48	6.11
Vector mean, D at °	1.35 at 97	1.52 at 180	2.33 at 1	2.27 at 154
TIA vector mean, D at °	1.35 at 7	1.52 at 90	2.33 at 91	2.27 at 64
6 Months
Achieved Vector mean, D at °	0.21 at 117	0.04 at 114	0.01 at 130	0.04 at 166
SIA vector mean, D at °	1.51 at 9*	1.55 at 90	2.33 at 91	2.23 at 64
DV vector mean, D at °	0.21 at 27†	0.04 at 24	0.01 at 40	0.04 at 76
Magnitude of error, D	0.04 ± 0.40	0.09 ± 0.49	−0.06 ± 0.56	−0.38 ± 1.50
Angle of error, ° ± SD	2 ± 11	0 ± 14	−2 ± 16	7 ± 22
Absolute angle of error, ° ± SD	6 ± 9	11 ± 9	12 ± 11	15 ± 17
Flattening effect, D ± SD	2.61 ± 1.43	1.83 ± 0.92	2.90 ± 1.50	4.34 ± 2.66
Flattening index	0.93	0.91	0.83	0.77
Correction index	0.99	1.03	0.96	0.99
Index of success	0.15	0.31	0.21	0.30
Residual astigmatism
Overall mean, D ± SD	0.39 ± 0.36	0.54 ± 0.45	0.64 ± 0.56	1.65 ± 1.37
At meridian of TIA, D ± SD	0.37 ± 0.59	0.41 ± 0.39	0.61 ± 0.55	1.37 ± 1.33
12 Months
Achieved Vector mean, D at °	0.18 at 100	0.06 at 108	0.20 at 95‡	0.01 at 60
SIA vector mean, D at °	1.53 at 7*	1.57 at 90	2.52 at 91	2.28 at 64
DV vector mean, D at °	0.18 at 10†	0.06 at 18	0.20 at 5	0.01 at 150
Magnitude of error, D ± SD	0.07 ± 0.40	0.21 ± 0.49‡	0.14 ± 0.54‡	−0.32 ± 1.51
Angle of error, ° ± SD	2 ± 11	1 ± 13	−2 ± 16	4 ± 21
Absolute angle of error, ° ± SD	5 ± 9	10 ± 9	12 ± 11	15 ± 16
Flattening effect, D ± SD	2.64 ± 1.42	1.97 ± 0.91	3.06 ± 1.52	4.40 ± 2.74
Flattening index	0.94	1.00‡	0.89‡	0.78
Correction index	1.00	1.11‡	1.04‡	0.99
Index of success	0.14	0.32	0.18	0.28
Residual astigmatism
Overall mean, D ± SD	0.37 ± 0.36	0.58 ± 0.54	0.59 ± 0.60	1.56 ± 1.41
At meridian of TIA, D ± SD	0.27 ± 0.31	0.44 ± 0.38	0.53 ± 0.63	1.40 ± 1.40

*  Significantly different from TIA (repeated measures ANOVA; P < 0.05).

†  Significantly different from zero (Hotelling Trace MANOVA; P < 0.05).

‡  Significantly different from 6 months (repeated measures ANOVA or paired t-test; P < 0.05.

Figure 1.

 

Double-angle vector diagrams of the preoperative and achieved refractive astigmatism outcomes at 12 months following AcrySof toric IOL (A), Artiflex toric pIOL (B), or Artisan toric pIOL implantation in normal eyes (C) or Artisan toric pIOL implantation in postkeratoplasty eyes (D). Data are shown as an ellipse of 1 SD with a center equal to the mean.

In the AcrySof toric IOL group, the SIA was significantly different from the TIA (P < 0.001). The DV was significantly different from zero at 6 months (0.21 D at 27°; P = 0.008) and 12 months (0.18 D at 10°; P = 0.043). Achieved astigmatism, ME, AE, flattening index, correction index, index of success, and residual astigmatism did not change from 6 to 12 months postoperatively (P > 0.05). 

In the Artiflex toric pIOL group, the SIA was not significantly different from the TIA (P > 0.05) and the DV at 6 and 12 months was not different from zero (P > 0.05). The ME (P = 0.026), flattening index (P = 0.013), and correction index (P = 0.029) changed significantly from 6 to 12 months follow-up, whereas achieved astigmatism, AE, index of success, and residual astigmatism did not change from 6 to 12 months follow-up (P > 0.05). 

In the Artisan pIOL group (normal eyes), the SIA at 6 and 12 months postoperatively was not significantly different from the TIA (P > 0.05). The DV at both 6 and 12 months was not significantly different from zero (P > 0.05). Achieved astigmatism (P = 0.020), ME (P = 0.013), flattening index (P = 0.020), and correction index (P = 0.006) showed a significant change from 6 to 12 months follow-up. The AE, index of success, and residual astigmatism did not change significantly from 6 to 12 months follow-up (P > 0.05). 

In the Artisan pIOL group (PKP eyes), the SIA was not significantly different from the TIA (P > 0.05) and the DV was not different from zero at 6 and 12 months (P > 0.05). Achieved astigmatism, ME, AE, flattening index, correction index, index of success, and residual astigmatism did not change from 6 to 12 months follow-up (P > 0.05). At 6 and 12 months postoperatively, all eyes had less than 3.0 D of refractive astigmatism. 

Predictor Variables of the Postoperative UDVA

Variables that were found to predict the UDVA at 6 and 12 months postoperatively are shown in Table 4. In normal eyes, the following variables were found to predict the UDVA at 12 months postoperatively: overall residual astigmatism (AcrySof group); overall residual astigmatism, preoperative CDVA, and ME (Artiflex group); and preoperative CDVA (Artisan group). None of the variables was found to predict the postoperative UDVA in PKP eyes. 

Table 4.

 

Predictor Variables of UDVA at 6 and 12 Months Postoperatively

	Toric IOL	Population	Predictor Variable Postoperative UDVA	Beta*	P Value
6 Months	AcrySof	Normal eyes	Index of success	0.489	0.010
	Artiflex	Normal eyes	Preoperative CDVA	0.675	<0.001
			Overall residual astigmatism	0.364	0.003
	Artisan	Normal eyes	Preoperative CDVA	0.705	<0.001
			Flattening index	0.296	0.007
	Artisan	PKP eyes	None
12 Months	AcrySof	Normal eyes	Overall residual astigmatism	0.749	<0.001
	Artiflex	Normal eyes	Overall residual astigmatism	0.578	<0.001
			Preoperative CDVA	0.382	0.006
			Magnitude of error	0.371	0.014
	Artisan	Normal eyes	Preoperative CDVA	0.776	<0.001
	Artisan	PKP eyes	None

*  Standardized coefficient of the multiple regression analysis.

Vector Analysis of Corneal Astigmatism

Postoperative changes in corneal astigmatism are shown in Table 5. Following a 2.2-mm superior incision, postoperative corneal astigmatism was not significantly different from the preoperative value (P > 0.05). In addition, the amount of SICA at 6 and 12 months postoperatively was not significantly different from zero (P > 0.05). Following a 3.4-mm and 5.4-mm superior incision in normal eyes, postoperative corneal astigmatism was significantly lower than the preoperative value (3.4 mm, P = 0.004; 5.4 mm, P = 0.001). The amount of SICA was significantly different from zero in both the 3.4-mm incision group (12 months: −0.33 D at 97°, P < 0.001) and the 5.4-mm incision group (12 months: −0.60 D at 90°, P <0.001). Following a 5.4-mm incision in PKP eyes, postoperative corneal astigmatism was not significantly different from preoperative astigmatism (P > 0.05). The SICA was −0.91 D at 76° (P = 0.018 compared to zero) at 12 months. The SICA did not change significantly from 6 to 12 months postoperatively in any of the incision groups. Figure 2 shows the mean amount of SICA with an ellipse of 1 SD for each incision group. 

Table 5.

 

SICA Following a 2.2-mm, 3.4-mm, or a 5.4-mm Incision in Normal Eyes, or Following a 5.4-mm Incision in PKP Eyes

Incision size, mm	2.2	3.4	5.4	5.4
Population	Normal eyes	Normal eyes	Normal eyes	PKP eyes
Preoperative corneal astigmatism
Arithmetic mean, D ± SD	2.62 ± 1.26	2.01 ± 0.70	3.13 ± 1.12	6.46 ± 2.34
Vector mean, D ± SD at degrees	1.21 ± 1.86 at 96	1.72 ± 0.88 at 91	2.34 ± 1.65 at 92	2.29 ± 4.54 at 68
Postoperative corneal astigmatism: 6 mo
Arithmetic mean, D ± SD	2.65 ± 1.11	1.79 ± 0.61	2.74 ± 1.00	5.87 ± 2.33
Vector mean, D ± SD at degrees	1.02 ± 1.86 at 90	1.43 ± 0.86 at 89*	1.86 ± 1.57 at 94*	1.68 ± 4.21 at 68
Postoperative corneal astigmatism: 12 mo
Arithmetic mean, D ± SD	2.70 ± 1.13	1.75 ± 0.68	2.57 ± 1.03	6.16 ± 2.73
Vector mean, D ± SD at degrees	1.12 ± 1.87 at 93	1.40 ± 0.87 at 90*	1.63 ± 1.54 at 93*	1.63 ± 4.44 at 62
SICA: 6 mo
Arithmetic mean, D ± SD	0.80 ± 0.72	0.63 ± 0.33	0.94 ± 0.50	2.03 ± 1.80
Vector mean, D ± SD at degrees	−0.25 ± 0.42 at 94	−0.31 ± 0.43 at 91†	−0.48 ± 0.55 at 91†	−0.49 ± 1.48 at 68†
SICA: 12 mo
Arithmetic mean, D ± SD	0.79 ± 0.74	0.77 ± 0.39	1.06 ± 0.75	2.58 ± 2.15
Vector mean, D ± SD at degrees	−0.10 ± 0.52 at 92	−0.33 ± 0.54 at 97†	−0.60 ± 0.53 at 90†	−0.91 ± 2.13 at 76†
P value‡	0.323	0.960	0.999	0.574

*  Significantly different from preoperative astigmatism (repeated measures ANOVA; P < 0.05).

†  Significantly different from zero (Hotelling Trace MANOVA; P < 0.05).

‡   P value of the difference between 6 and 12 months, within a group (repeated measures ANOVA).

Figure 2.

 

Double-angle vector diagrams of the surgically induced corneal astigmatism at 12 months following a 2.2-mm (A), 3.4-mm (B), and a 5.4-mm incision in normal eyes (C) or a 5.4-mm incision in postkeratoplasty eyes (D). Data presented as mean with an ellipse of 1 SD.

Complications

In the AcrySof toric IOL group, one patient developed postoperative cystoid macular edema, which was treated with ketorolac (Acular; Allergan, Irvine, CA). In 4 eyes in the Artiflex toric pIOL group, postoperative depositions were seen on the anterior and/or posterior surface of the pIOL. Two eyes had a postoperative elevated intraocular pressure (IOP), which was treated with timolol (Timoptol XE; Merck Sharp & Dohme, Haarlem, The Netherlands). In the Artisan toric IOL group (normal eyes), one patient had a postoperative elevated IOP, which was treated with timolol. In the Artisan toric IOL group (PKP eyes), a pIOL claw repositioning was performed at 6 weeks postoperatively. One patient developed a high IOP, which was treated with timolol. 

Discussion

The aim of this study was to determine the changes in refractive and corneal astigmatism following toric IOL and toric pIOL implantation. With a vector analysis based on refractive astigmatism, the effectiveness of the astigmatism correction was determined. In addition, the amount of SICA following a 2.2-mm, 3.4-mm, and 5.4-mm incision was determined at 6 and 12 months postoperatively. Even though the visual outcomes following toric IOL and toric pIOL implantation were good, the postoperative CDVA was significantly better than the UDVA in all subgroups. It is our belief that more knowledge about the amount of SICA can further increase the effectiveness of toric IOLs. 

To determine the effectiveness of the astigmatic correction with different toric IOLs and toric pIOLs, a vector analysis according to Alpins was performed. Alpins astigmatism analysis is based on 3 fundamental vectors: the TIA, SIA, and DV. ¹³ These three vectors are then used to calculate different astigmatic parameters and indices that may be used to determine overall success of the astigmatism correction (index of success, flattening index), possible over- or undercorrection (ME, correction index), misalignment of treatment (AE, absolute AE), and residual astigmatism (FE, residual astigmatism). 

Following AcrySof toric IOL implantation, the ME was close to zero and the correction index was 1.00, indicating that on average no over- or undercorrection of astigmatism had occurred. Previous studies that have performed a vector analysis following toric IOL implantation report either a slight overcorrection ^16,17 or an undercorrection ^1,18 of astigmatism. In addition, a recent study by Goggin et al. ¹⁹ shows that the manufacturer underestimated the IOL cylinder power at the corneal plane, which may result in an overcorrection of astigmatism. In our study the absolute AE was 5 ± 9°, which is in accordance with the mean misalignment reported in previous studies. ^2,17,18,20 At 12 months postoperatively, the overall mean magnitude of residual astigmatism was 0.37 ± 0.36 D, and 0.27 ± 0.31 D of astigmatism remained at the meridian of TIA. As discussed in our previous study, two views currently exist in the literature regarding the calculation of residual astigmatism: the overall magnitude of residual astigmatism and the magnitude of residual astigmatism at the meridian of treatment. ¹⁴ Our results indicate that both methods result in similar amounts of residual astigmatism, although the amount of residual astigmatism at the meridian of treatment was always slightly lower than the overall mean. At 12 months postoperatively, the overall magnitude of residual astigmatism was found to predict the postoperative UDVA. None of the measured astigmatism parameters, or the UDVA or CDVA, changed significantly from 6 to 12 months follow-up, indicating that visual and refractive outcomes were stable at 6 months follow-up. 

In the Artiflex and Artisan pIOL groups, the ME and correction index showed an overcorrection of astigmatism The mean absolute AE was 10 ± 9° and 12 ± 11° in the Ariflex and Artisan pIOL groups, respectively. This indicates a slightly higher misalignment of treatment in toric pIOLs than in toric pseudophakic IOLs. Residual astigmatism with both pIOLs was comparable: 0.58 to 0.59 D overall and 0.44 to 0.53 D at the meridian of treatment. The preoperative CDVA, overall magnitude of residual astigmatism, and ME were found to predict the UDVA at 12 months postoperatively. In both pIOL groups, astigmatism parameters such as the ME, flattening index, and correction index changed from 6 to 12 months postoperatively, indicating that the refractive outcomes are not entirely stable at 6 months. However, visual outcomes did not change between 6 and 12 months follow-up. 

Following Artisan toric pIOL implantation in PKP eyes, the variability in the achieved astigmatic outcomes was much higher than that in normal eyes (demonstrated in Figure 1). At 12 months follow-up, the ME was −0.32 ± 1.51 D and the correction index was 0.99, indicating that an undercorrection of astigmatism occurred in these patients. The high mean absolute AE (15 ± 16°) resulted in large amounts of residual astigmatism (1.56 ± 1.41 D overall and 1.40 ± 1.40 D at the meridian of treatment). None of the astigmatism parameters changed significantly from 6 to 12 months follow-up. All eyes in our study achieved less than 3.0 D of refractive astigmatism, indicating that all patients were suitable for spectacle correction. 

An important aspect to consider in toric IOL calculation is the vector change in corneal astigmatism (SICA) induced by the incision. Commonly used toric IOLs require an incision of 2.2 mm (AcrySof toric IOLs) or can even be implanted through a sub 2.0-mm incision (T-flex toric IOLs, Rayner, East Sussex, UK; AT Torbi, Carl Zeiss Meditec, Jena, Germany). Following a 2.2-mm superior incision, a SICA of −0.10 ± 0.52 D at 92° was found at 12 months postoperatively, which was not significantly different from zero. In addition, postoperative corneal astigmatism was not significantly different from preoperative corneal astigmatism. Studies on toric IOLs generally incorporate a SICA, ranging in magnitude from 0 to 0.70 D. ^{18,19,21–24} In addition, previous studies have reported a SICA ranging from −0.20 D to −0.40 D following a 2.2-mm superior incision. ^4,25–27 However, none of these aforementioned studies have examined if the amount of SICA is significantly different from zero. In agreement with our results, Hoffmann et al. ¹⁶ have reported a SICA of −0.19 ± 0.44 D at 3 months postoperatively, following a 2.2-mm temporal incision, which was not significantly different from zero. In addition, Goggin et al. ²⁸ have examined the test–retest variability of keratometry measurements and have shown that this may be up to 0.14 D in magnitude. This indicates that a significant proportion of the SICA reported in previous studies may be due to test–retest variation in keratometry measurements. It is our belief that no SICA should be incorporated into the toric IOL power calculation for a superior 2.2-mm incision. 

Anterior chamber iris-fixated toric pIOLs require a 3.4-mm incision for foldable pIOLs or a 5.4-mm incision for rigid pIOLs. A SICA of −0.31 ± 0.43 D for a 3.4-mm superior incision and −0.48 ± 0.55 D for a 5.4-mm superior incision was found at 6 months follow-up in normal eyes. This did not change significantly from 6 to 12 months postoperatively, indicating that the amount of SICA had stabilized at 6 months postoperatively. In both incision groups, postoperative corneal astigmatism was significantly different from the preoperative value and the amount of SIA was significantly different from zero. This indicates that the amount of SICA should be incorporated into the pIOL power calculation. However, the amount of SICA is currently not routinely incorporated into the pIOL power calculation performed by the manufacturer. Not incorporating the SICA into the pIOL power calculation may have resulted in the overcorrection of astigmatism observed in our study. In the literature, superior incisions of approximately 3.5 mm have been reported to produce a SICA ranging from −0.49 to −0.89 D. ^29–31 Bartels et al. ^32,33 have examined the SICA, following a 5.5-mm incision, and have shown a SICA ranging from −0.60 to 0.74 D. Our recommendation is to incorporate a SICA of −0.30 D for a 3.4-mm superior incision and −0.50 D for a 5.4-mm superior incision. 

In PKP eyes, the SICA, following a 5.4-mm incision, was found to be −0.49 ± 1.48 D, which is comparable to the SICA in normal eyes. However, as is visible in Figure 2, the variability in the amount of SICA in PKP eyes was much higher than in normal eyes. It was hypothesized that the large variability in SICA might be due to a different biomechanical response of the cornea. In normal corneas, collagen fibers within the same lamellae have a parallel orientation, and the orientation of the fibers throughout the corneal depth varies. Overall, the preferred orientation is along the horizontal (nasal–temporal) and vertical (inferior–superior) direction, which provides biomechanical strength. ³⁴ In PKP eyes, however, collagen fibers are cut and the donor cornea button is sutured into place with no regard to the orientation of the recipient and donor collagen lamellae. This might impair the structural integrity of the cornea and change the biomechanical response. Since the amount of SICA in PKP eyes in this study was found to be significantly different from zero, incorporation of a SICA of −0.50 D into the toric pIOL power calculation, for a superior 5.4-mm incision, is advised. 

A limitation of this study is that preoperative and postoperative visual acuity was measured with a Snellen chart instead of an ETDRS (LogMAR) chart. However, Snellen visual acuity was converted to LogMAR before any calculations were performed. A possible limitation is that, for some patients, both eyes were included in the analysis. Therefore, all analyses were also performed with only one eye for each patient (results not shown). These results were similar to the results for all eyes. A third limitation is that all surgeries were performed by two surgeons. Another limitation is that two different devices were used for corneal topography. To attempt to minimize this effect, the same device was used preoperatively and postoperatively in each patient. Finally, a limitation of this study is that corneal astigmatism data were obtained with Placido disk videokeratoscopy. This type of corneal topography system reconstructs the anterior corneal surface by the reflections of light-emitting Placido rings. This does not represent the exact corneal shape, since it does not include information about the posterior corneal surface and the corneal thickness. ³⁵ It would be worthwhile in the future to analyze astigmatism data in these patients with Scheimpflug imaging. 

In conclusion, a vector analysis of the refractive astigmatism changes showed that toric IOLs and toric pIOLs provide an effective astigmatism correction. In the case of toric IOL implantation through a superior 2.2-mm incision, incorporation of a SICA into the toric IOL power calculation is not recommended. In the case of Artiflex or Artisan pIOL implantation through a superior 3.4- or 5.4-mm incision, incorporation of a SICA of −0.30 D and −0.50 D, respectively, is recommended. In the case of Artisan pIOL implantation in PKP eyes, incorporation of a SICA of −0.50 D is also recommended. Incorporating the SICA into the pIOL power calculation may further increase the effectiveness of the astigmatism correction. 

Footnotes

 Disclosure: N. Visser, None; T.T.J.M. Berendschot, None; N.J.C. Bauer, None; R.M.M.A. Nuijts, None