It is well recognized that even small amounts of refractive astigmatism can have a significant effect on visual acuity (VA)1; therefore, precise elimination of astigmatism at the time of laser in situ keratomileusis (LASIK) is highly desirable. The challenge for refractive surgeons is that the preoperative refractive astigmatism measured in a phoropter may be affected by higher-order aberrations of the eye, notably coma.2
Currently, excimer laser systems designed to perform LASIK can correct astigmatism in different ways. The most basic method is to treat the measured refractive astigmatism based on the manifest refraction. A more advanced alternative is to use a wavefront-guided ablation that requires measuring the ocular wavefront of the entire eye and designing a specific ablation pattern to reduce or eliminate the ocular aberrations measured. Results with this approach appear only modestly better than treating the measured refractive astigmatism, when better at all.3 It is possible that the ability to accurately measure the ocular wavefront is still insufficient to provide a consistent beneficial effect. A third alternative is topography-guided LASIK that attempts to create a more uniform corneal surface to reduce aberrations while also correcting the refractive error. This technology was initially reserved to correct irregular astigmatism,4 although recent advances have led to it being used on virgin corneas as well, with promising results.
One topography-guided LASIK procedure (Contoura Vision, Alcon) has been shown to provide excellent visual outcomes in normal eyes without irregular astigmatism, with 93% of eyes achieving 20/20 or better uncorrected distance VA (UDVA), 65% achieving 20/16 or better UDVA, and 34% achieving 20/12.5 or better UDVA 1 year postoperatively.5 These results were achieved in a trial designed to obtain U.S. Food and Drug Administration (FDA) approval for the Contoura. There were limitations to the trial design, as the clinical protocol required that investigators enter the manifest refraction as the treatment refraction. In addition, there was a very close agreement between the manifest and topographic cylinder magnitude and axis because subjects with a poor agreement were specifically excluded from the trial.
After FDA approval of the topography-guided Contoura algorithm, it became more widely used in clinical settings. Treatments were performed on eyes with a greater disparity between the manifest and topographic cylinder measurements. The challenge has been determining the best balance between using the topographic information and the refractive information to provide an optimal correction for each eye. The topography-guided treatment profile is fixed by the Contoura software, designed to treat all corneal topographic aberrations. However, the sphere and cylinder correction are user specified; the WaveLight Topolyzer VARIO (Alcon) laser provides the user with the opportunity to select a treatment of sphere and cylinder. Different surgeons have based the sphere and cylinder treatment on the desired clinical outcome, the manifest refraction, the topographic cylinder, and/or other factors. Subjective approaches vary by surgeon, although more objective algorithms for treatment planning, such as that of Kanellopoulos,6 have been developed. These rely on either treating completely from the measured topographic anterior astigmatic magnitude and axis or systematically combining the elements of both manifest refractive and topographic cylinder measurements. Others have suggested that using the manifest refractive cylinder alone as the basis for topography-guided treatment remains the best option.7
A new alternative to the above-noted approaches, the Phorcides Analytic Engine (Phorcides LLC), considers the topography of the cornea in terms of not simply astigmatism but also its varying elevation profile. As such, it is not constrained by using only corneal astigmatism magnitude and axis as an input to the ablation algorithm. It may, therefore, provide an ablation pattern that is more responsive to asymmetry on the cornea. Corneal asymmetry is often the source of higher-order aberrations such as coma. A patient's manifest refraction may include a cylinder magnitude that is selected to reduce the blur circle created by coma. If the coma is effectively treated with topography-guided LASIK, the lower-order cylinder correction needed would likely be different.2 Another consideration is that posterior corneal astigmatism and lenticular astigmatism affect the path of light through the eye. The Phorcides Analytic Engine thus considers the anterior corneal astigmatism, the topographic irregularities that create higher-order aberrations, the posterior corneal astigmatism, and the lenticular astigmatism when determining the optimal treatment of an eye with topography-guided LASIK.
The purpose of this retrospective study was to compare the postoperative results (2 to 5 months postoperatively) achieved in eyes treated with Contoura based on the Phorcides Analytic Engine with the results achieved when the manifest refraction was the basis for Contoura treatment planning.
This study was a double-arm retrospective chart review and a comparative noninterventional study of postoperative refraction and VA at least 2 months after uneventful LASIK using the Contoura algorithm on the WaveLight laser. The study was approved by an institutional review board (SALUS IRB, Austin, TX). Clinical results from 4 surgeons (K.S., T.T., S.W., M.L.) at 4 sites were extracted from clinical charts. No protected health information was recorded.
Eligible eyes were those from patients who had previous LASIK using Contoura, with treatments in the approved range: up to −9.0 diopters (D) of spherical equivalent myopia or myopia with astigmatism, with up to −8.0 D of the spherical component and up to 3.0 D of the astigmatic component at the spectacle plane. Both eyes were eligible for analysis. Surgical planning had to be based on either the manifest refraction (manifest group) or the Phorcides Analytic Engine (analytic group), with nomogram adjustments permitted in both groups. The manifest refraction and VA (corrected and uncorrected) from 2 to 3 months postoperatively had to be available. To avoid confounding factors, eyes with clinically significant ocular pathology (eg, keratoconus and basement membrane dystrophy) other than residual refractive errors or irregular astigmatism were excluded, as were eyes with suboptimal surgical outcomes that were not related to the treatment profile chosen (eg, flap dislocation). Eyes treated for monovision were also excluded. Each surgeon was asked to tabulate 80 eyes treated with the Phorcides planning software and 80 eyes that were treated using the manifest refraction.
Manual and electronic data records were used to identify eyes that fit the inclusion and exclusion criteria. Deidentified data from the preoperative examination (age, sex, VA, and refraction), the treatment plan (keratometry and nomogram-adjusted treatment parameters), and the postoperative examination 2 to 5 months postoperatively (VA and refraction) were collected; the source of the postoperative data (internal or from comanaging physicians) was also recorded. Where possible, any recorded enhancement in the first year after initial surgery was also recorded, although no eyes having undergone an enhancement before the recorded postoperative visit were included in the study pool. Outcomes of interest were the postoperative manifest refraction, the distribution of residual refractive astigmatism, and VA (uncorrected and corrected). All VA data were converted to logarithm of the minimum angle of resolution (logMAR) for analysis. Correlations between the clinical results and the vector difference between the preoperative topographic and manifest cylinder were also investigated.
Statistical analyses were performed using the Statistica software (version 12, TIBCO Software Inc.). Categorical data were tested using appropriate nonparametric tests (eg, χ2 test), whereas continuous variables were tested using an analysis of variance. All statistical tests of hypotheses were based on a level of significance of α = 0.05.
Data from 640 eyes were extracted from the clinical records: 317 eyes treated with the manifest refraction and 323 eyes treated using the analytical engine. Table 1 contains the baseline demographic and treatment information. The groups were well matched for age, sex, and refractive/topography status, with the only significant difference being the number of patients comanaged; more patients in the manifest group were comanaged. Follow-up times varied from 59 to 155 days (2 to 5 months), with 541 (85%) of 640 eyes seen between 2 and 3 months postoperatively. Table 2 shows the breakdown by surgeon; differences in the preoperative data were not clinically significant, so the data were pooled for analysis. All surgeons contributed at least 70 eyes to each surgical group.
As a retrospective analysis, refractive outcomes data were collected for clinical, rather than research, purposes; in many instances, the VA was recorded by line, rather than by letter. This produced some atypical results. For instance, there were 138 cases (21%) in which a manifest refraction was not plano, but the measured corrected distance VA (CDVA) was the same as the UDVA. However, the manifest refractions for these cases were generally within ±0.25 D of plano with less than 0.5 D of the cylinder.
Figure 1 shows the mean postoperative spherical equivalent refraction (SEQ) by surgeon and group. There were statistically significant differences by surgeon (P = .01) and by group (P < .01), but no interaction effect. The mean differences were less than 0.07 D, which were considered clinically insignificant. The percentage of eyes within ±0.25 D of plano was not statistically significantly different between groups (269/317 [84.9%] in the manifest group vs 263/323 [81.4%] in the analytic group; P = .25).
Figure 2 shows a histogram of the postoperative refractive cylinder by group. The mean residual cylinder was 0.15 ± 0.33 D in the manifest group and 0.16 ± 0.32 D in the analytic group. There was no statistically significant difference between groups (P = .79, analysis of variance); 96% of eyes in both groups showed a refractive cylinder of less than 0.50 D postoperatively. Only 6 eyes (4 manifest and 2 analytic) exhibited a postoperative refractive cylinder of 1.0 D or greater.
Figure 3 shows a histogram of the postoperative UDVA by group. Over 94% of eyes treated had 20/20 or better UDVA postoperatively in both groups, but significantly more eyes had 20/16 or better (−0.1 logMAR) UDVA in the analytic group (131/317 [41.3%] in the manifest group vs 202/323 [62.5%] in the analytic group; P < .001, χ2 test). Although not shown, the percentage of eyes with 20/16 or better postoperative CDVA was also statistically significantly higher in the analytic group (163/317 [51.4%] in the manifest group vs 249/323 [77.1%] in the analytic group; P < .001, χ2 test). All eyes but 1 had a CDVA of 20/20 or better postoperatively.
Figure 4 shows how the postoperative UDVA compared with the patient's preoperative CDVA. As can be seen, most patients (78% of manifest and 85% of analytic) showed a postoperative UDVA equal to or better than their preoperative CDVA. The number of patients with a UDVA better than their preoperative CDVA was significantly higher in the analytic group (73/317 [23.0%] in the manifest group vs 118/323 [36.5%] in the analytic group; P < .001, χ2 test). Figure 5 shows a similar comparison between the postoperative CDVA and preoperative CDVA. No eye in either group lost more than 1 line of CDVA. Significantly more eyes in the analytic group gained 1 or more lines of CDVA (96/317 [30.3%] in the manifest group vs 138/323 [42.7%] in the analytic group; P = .001, χ2 test).
The vector difference between the preoperative topographic and manifest cylinder was calculated. Correlations between the magnitude of this vector and the clinical outcomes were calculated. The only statistically significant correlation was with the postoperative UDVA in the manifest group, but the correlation was weak (r = 0.15). A second test was made that categorized the eyes in terms of whether they would have met the criteria for inclusion into the study conducted to obtain FDA approval. Eyes in the FDA trial were only included if 3 separate corneal specialists found a normal corneal topography and tight agreement between the manifest astigmatism and the measured anterior corneal astigmatism. To try to match these criteria for use by surgeons in clinical practice, Alcon has suggested that eyes in which the cylinder magnitude difference between Contoura and manifest ≤0.75 D and the absolute angle difference within 5 degrees if the manifest cylinder is ≥2.0 D or within 10 degrees if the manifest cylinder is ≤1.75 D most closely match those used in the FDA TCAT (topography-guided customized ablation treatment; THINK Surgical, Inc.) study. Of the 640 eyes in the dataset, 153 (23.9%) met the TCAT study criteria. There was no significant effect of this categorization with regard to the postoperative manifest refraction spherical equivalent (P = .59) CDVA (P = .10) or UDVA (P = .054). However, there was a statistically significant effect with regard to the postoperative cylinder, but the difference was not clinically significant (manifest 0.15 ± 0.22, analytic 0.20 ± 0.23; P = 0.01).
Using the manifest refractive astigmatism for surgical planning was adopted in Contoura based on the requirements for the FDA study. A recently conducted large retrospective study suggests that using the manifest refractive astigmatism axis as opposed to the topography-based anterior corneal astigmatism axis resulted in better outcomes for patients with significant astigmatism and a significant difference in the 2 axes.7 Results from several studies suggest that topography-guided LASIK induces less postoperative aberration in treated eyes.8,9 However, Zhang and Chen10 reported a greater residual refractive astigmatism with topography-guided treatment when compared with wavefront-optimized treatment; they suggested an improved astigmatism algorithm may be necessary for topography-guided LASIK. Kanellopoulos noted that there is a likely benefit in modifying the astigmatic refraction for use in topography-guided LASIK, noting a 16% increase in the number of eyes with less than 0.50 D of residual astigmatism when compared with the manifest refraction.6
The goal of this study was to determine whether using the Phorcides Analytic Engine algorithm with Contoura LASIK improved clinical outcomes when compared with using the manifest refraction for treatment planning. It was a retrospective evaluation of 317 eyes treated with the manifest refraction and 323 treated using Analytic from 4 clinical sites. The results showed similar SEQ between the 2 groups; the percentage of eyes within ±0.25 D of plano was not statistically significantly different between the groups. The mean SEQ recorded in the literature was also between −0.06 D11 and 0.06 D5 at 3 months; this is reasonably consistent with results shown in Figure 1. Wallerstein et al.7 noted that the percentage of eyes achieving an SEQ within ±0.25 D was 80.0% when using manifest refraction, which is also reasonably consistent with this study.
The mean postoperative cylinder in this study was 0.16 D; this is consistent with results of 0.19 D reported by Moshirfar et al.12 The percentage of eyes with residual cylinder ≤0.50 D was 96% in both groups in this study. Tan et al.13 evaluated 2009 eyes treated with topography-guided LASIK and noted that 98% of these eyes showed ≤0.50 D of the residual cylinder; this is reasonably consistent with the results in this study. This study results may have been slightly more variable because the data were retrospective and included refractions from both surgeons and comanaging doctors.
The original FDA study with Contoura excluded eyes with a large difference between the manifest and topographic astigmatism. Of the 640 eyes in this study, only 24% would have met the FDA Contoura study criteria. The group here is likely more representative of the population that would be seen in a typical clinical experience, suggesting that 3 of 4 patients presenting for surgery were not represented in the FDA study. As such, it was of interest to note whether the difference in the vector between the preoperative topographic and manifest cylinder correlated with the clinical outcomes. No strong correlations were noted for any of the measured outcomes.
In the FDA study, with its relatively restrictive inclusion criteria, 247 eyes of 212 patients had refractive and VA data available 3 months postoperatively; 59% of eyes had 20/16 or better UDVA, and 88% had 20/20 or better UDVA.5 These results are slightly lower than the results in this study using the Phorcides Analytic Engine (20/16 or better [63%] and 20/20 or better UDVA [94%]), despite the inclusion of eyes with much larger differences between topographic and refractive astigmatism in this study. The results in this study for both UDVA and CDVA are also slightly better than those reported by Kanellopoulos in his randomized prospective study of a topographic manifest refraction LASIK algorithm.6 Kanellopoulos noted that 61% had a CDVA of 20/16 or better and 50% had a UDVA of 20/16 or better6; this compares with 77% in the analytic group with a CDVA of 20/16 or better and 63% with a UDVA of 20/16 or better in this study.
The percentage of patients with a UDVA better than their preoperative CDVA was 37% in the analytic group and 23% in the manifest group. Other studies evaluating results with topography-guided LASIK noted that up to 28% gained 1 line,6 26% gained 1 or more lines,5 and 11% gained 2 or more lines.6
There are limitations to this study, with the most important one being that it was a retrospective data review. Detecting small refractive and VA differences is likely to be more reliable when prospective evaluations are conducted with a consistent technique and with the use of logMAR charts to measure the VA to the letter. More precise refractive data would also allow for the determination of why the analytic and manifest groups differ in the VA results but not in the magnitude of the refractive cylinder. This may be due to changes in the preoperative to postoperative cylinder axis, but the retrospective data here do not lend themselves to such detailed analysis. Masking postoperative clinicians would also be a possible option for future studies. Plans are being prepared to conduct a prospective study to address these limitations.
In conclusion, this study demonstrates that using the Phorcides Analytic Engine for topography-guided surgery planning increased the likelihood of 20/16 UDVA and CDVA relative to using the manifest refraction.
WHAT WAS KNOWN
- The Contoura algorithm has been shown to provide a high percentage of eyes with 20/16 or better visual acuity (VA) postoperatively.
WHAT THIS PAPER ADDS
- The magnitude of residual refractive errors (sphere and cylinder) was not significantly different between the manifest refraction and Phorcides Analytic Engine methods of surgical planning.
- The percentage of eyes with 20/16 or better uncorrected and corrected distance VA postoperatively was significantly higher in the analytic engine group.
1. Atchison DA, Mathur A. Visual acuity with astigmatic blur. Optom Vis Sci 2011;88:E798–E805
2. Zhou W, Stojanovic A, Utheim TP. Assessment of refractive astigmatism and simulated therapeutic refractive surgery strategies in coma-like-aberrations-dominant corneal optics. Eye Vis (Lond) 2016;3:13
3. Manche E, Roe J. Recent advances in wavefront-guided LASIK. Curr Opin Ophthalmol 2018;29:286–291
4. Holland S, Lin DT, Tan JC. Topography-guided laser refractive surgery. Curr Opin Ophthalmol 2013;24:302–309
5. Stulting RD, Fant BS; T-CAT Study Group, Bond W, Chotiner B, Durrie D, Gordon M, Milauskas A, Moore C, Slade S, Randleman JB, Stonecipher K. Results of topography-guided laser in situ keratomileusis custom ablation treatment with a refractive excimer laser. J Cataract Refract Surg 2016;42:11–18
6. Kanellopoulos AJ. Topography-modified refraction (TMR): adjustment of treated cylinder amount and axis to the topography versus standard clinical refraction in myopic topography-guided LASIK. Clin Ophthalmol 2016;10:2213–2221
7. Wallerstein A, Gauvin M, Qi SR, Bashour M, Cohen M. Primary topography-guided LASIK: treating manifest refractive astigmatism versus topography-measured anterior corneal astigmatism. J Refract Surg 2019;35:15–23
8. Ozulken K, Yuksel E, Tekin K, Kiziltoprak H, Aydogan S. Comparison of wavefront-optimized ablation and topography-guided Contoura ablation with LYRA protocol in LASIK. J Refract Surg 2019;35:222–229
9. Tiwari NN, Sachdev GS, Ramamurthy S, Dandapani R. Comparative analysis of visual outcomes and ocular aberrations following wavefront optimized and topography-guided customized femtosecond laser in situ keratomileusis for myopia and myopic astigmatism: a contralateral eye study. Indian J Ophthalmol 2018;66:1558–1561
10. Zhang Y, Chen Y. A randomized comparative study of topography-guided versus wavefront-optimized FS-LASIK for correcting myopia and myopic astigmatism. J Refract Surg 2019;35:575–582
11. Waring G, Dougherty PJ, Chayet A, Fischer J, Fant B, Stevens G, Bains HS. Topographically guided LASIK for myopia using the Nidek CXII customized aspheric treatment zone (CATz). Trans Am Ophthalmol Soc 2007;105:240–248
12. Moshirfar M, Shah TJ, Skanchy DF, Linn SH, Kang P, Durrie DS. Comparison and analysis of FDA reported visual outcomes of the three latest platforms for LASIK: wavefront guided Visx iDesign, topography guided WaveLight Allegro Contoura, and topography guided Nidek EC-5000 CATz. Clin Ophthalmol 2017;11:135–147
13. Tan J, Simon D, Mrochen M, Por YM. Clinical results of topography-based customized ablations for myopia and myopic astigmatism. J Refract Surg 2012;28:S829–S836