In his Letter to the Editor (October), Dr. David Shapiro takes exception to my statement, "don't tell patients pupil size doesn't matter." (From "Screening for LASIK: Tips and Techniques" July 2006). In fact the headline over his letter [not written by Dr. Shapiro] states "Too Cautious Counsel on Pupil Size and LASIK?" He states the article "cautions against surgery in large pupil patients." In fact, I stated that I have operated on many of these patients utilizing wavefront-based surgery, including a patient with 9-mm pupils. Even though my personal preference is for sequential surgery in all patients, all of my large pupil patients have elected to have surgery in their second eye, even when they notice an increase in night vision complaints (NVC) compared to their preop vision. Rather than cautioning against the surgery, I am recommending a discussion of this potential issue with patients who have unusually large pupils. This raises the philosophical question: Should we as refractive surgeons, recommending an elective procedure to reduce the need for glasses, be too cautious or less cautious? Dr. Shapiro evidently feels we can be less cautious in this regard and uses his reading of the literature to support his position.
He talks about the large, excellent studies of Schallhorn and Pop. Let's take a close look at the "large" Schallhorn study. The article by Schallhorn1 in 2003 on 100 patients with 6-mm ablation zones with refractive errors from -2.88 to -9.25 D found that patients with large mesopic pupils did have more quality of vision issues at one month, but not at six months. Of the 100 eyes studied, only two were measured as greater than 7 mm under mesopic conditions, no doubt an underestimate of the true number of patients with scotopic pupils larger than 7 mm. Why do I say this? The pupils were measured under low light (<5 lux) with the Colvard device.
An article by Netto defines scotopic measurements using the Procyon pupillometer as being performed at 0.04 lux and mesopic high at 4 lux.2 That group studied 192 eyes and found that the mean scotopic pupil size was about 6.5 mm; the mean low mesopic pupil size was about 5.6 mm; and the mean high mesopic was about 4.0 mm. A study by Twa compared infrared video and digital photography to the Colvard at three different lux settings.3 On average the infrared or digital measurements at <0.63 lux (scotopic) were about 0.5-mm larger than the measurements with the Colvard at about 5 lux. Because of the testing method, the Schallhorn study most likely underestimated true scotopic pupil size. If just one or two patients with "small pupils" who had NVC actually had scotopic pupils greater than 6.5 mm, the analysis and conclusions would have been different. The correlation with NVC and large pupils present at one month may have also been present at six months. Since one-third of these 100 patients missed their last follow-up visit, it is uncertain whether the conclusion that the correlation with pupil size present at one month was not still true at six months. The fact that many patients with NVC complaints and large pupils "cortically adapt" to their new vision doesn't mean the pupils were not a factor in their original complaints.
The other recent study which failed to show a correlation with pupil size and NVC was the study by Pop.4 I would agree that this is truly a large study of 795 patients with myopia up to -9.75 D. This study found that pupil size was not predictive of NVC at any postoperative month. Attempted correction, optical zone and residual refractive error were the major risk factors for NVC. Once again the conclusion about pupil size may not be valid, as the Colvard instrument was the measuring device. Although it was stated the measurements were "scotopic," the testing situation in terms of lux level was not stated, so these measurements may really have been mesopic. It is also unlikely that the patients had 10 minutes of dark adaptation prior to measurement emphasized by Brown as important to obtain true scotopic pupil size.5 I believe that Dr. Pop was able to customize the ablation zone and blend zone with the Nidek laser (possibly using a larger ablation diameter and blend zone in patients with larger pupils) in a manner not available in the United States, so applying this study to current U.S. systems may not be valid.
In 1993, Roberts used an optical analysis computer program to study the effect of the optical zone with entrance pupils of 2 to 8 mm.6 The conclusion was "Optical zone diameter must be at least as large as the entrance pupil diameter to preclude glare at the fovea, and larger than the entrance pupil to preclude parafoveal glare." This is, of course, not always possible with high corrections and large pupils because of ablation depth constraints.
In 1996, Klonos and Pallikaris used a computer model to predict image quality after photorefractive keratectomy.7 In this article they make a statement similar to the conclusion by Roberts, which seems logical to us: "When the ablation zone covers the entire entrance pupil, the image quality may be uniform. However, if the entrance pupil is not fully covered by the ablation zone, disturbances in night vision often are reported as starbursts and halos around bright sources of light." In discussing their "Excimeye" ray tracing modeling program they state: "For example, lateral spherical aberration is the radius of the retinal blur circle relative to pupil diameter and, therefore, the wider the pupil size is, the bigger the blur circle is, resulting in a bad retinal image."
When scotopic pupil size is greater than 7 mm, the ablation diameter cannot usually be made larger than the pupil size as recommended by Roberts and Klonos, so one would expect patients with these larger pupil sizes who have laser vision surgery to have more NVC. The fact that many of them eventually adapt to their situation, as in the Schallhorn study, should not lead us to conclude that their large scotopic pupil size is irrelevant and not a risk factor.
Wavefront measurement of higher-order aberrations gives us a new tool to objectively evaluate quality of vision issues following laser vision correction. It is an accepted fact that the measurement of higher-order aberrations is related to the aperture size of the measurement. The larger the pupil size, the greater the higher order aberration scores. Maguen et al showed that there was a statistically significant increase in total aberrations in ametropic eyes as the pupil size increased in both pre- and postoperative LASIK patients.8,9 A study by Randazzo on management of night vision disturbances showed a progressive decrease in higher-order aberrations as well as a reduction in glare and halos as the pupil diameter decreased.10
This relationship and the correlation with NVC are evident in a recent study by Chalita and Krueger on 105 post LASIK patients studied retrospectively with the Alcon LadarWave device.11 In this study, scotopic pupil size ranged from 3.0 mm to 8.5 mm: "The larger the pupil size evaluated, the higher the ocular aberration values." When they compared the higher order aberration for 5- and 7-mm apertures, they found that "there is a significant difference between them (p<0.001) showing that the larger apertures have more aberrations in post-LASIK eyes." They found a positive correlation between scotopic pupil size and starbursts (p=0.001). Glare complaints were significantly associated with spherical aberration and total aberration, and both of these were higher with increasing scotopic pupil size.
Helgeson found large pupil size measured preoperatively to correlate with an increased frequency of visual disturbances in post-LASIK patients under scotopic conditions.12
We acknowledge that the cause of NVC's is multifactorial, and certainly amount of correction, ablation zone, residual refractive error, corneal haze and decentration all play a role. If you tell patients with large pupils that they are not at greater risk for NVCs than patients with smaller pupils, and they in fact experience significant NVCs, you will be hard-pressed to explain why their symptoms are reduced when their pupils are made smaller. This has, in fact, been our experience with large pupil patients who have NVCs. Their symptoms and their higher-order aberration scores usually diminish when their pupils are reduced. Patients with average to smaller scotopic pupils usually do not experience relief from a reduction in pupil size.
Both the Visx13 and Alcon14 patient information brochures on wavefront-based treatments mention the possible increased risk of treating patients with large pupils. Dr. Shapiro mentioned the Visx CustomVue trial several times in his article. Does he ignore the following statement from the Visx CustomVue patient information booklet (Facts you need to know about CustomVue. 2003): "Larger pupils. Before surgery, your doctor should measure your pupil size under dim light conditions. You might have difficulty seeing in dim lighting, rain, snow, or bright glare. Whether you may have poor vision under these conditions is hard to predict because it has been studied so little." The Alcon brochure for CustomCornea states: "Before surgery, your doctor should measure your pupil under dim lighting conditions. If your pupils in dim light are >6.5mm, consult with your doctor about the risk that the surgery may cause negative effects on your vision, such as glare, halos, and night driving difficulty." The FDA expects these brochures to be given to each patient.
Dr. Shapiro stated that the literature does not support that "larger pupils are associated with greater risk of night vision problems after LASIK." We have tried to show that there is literature that does support this correlation. I would assume by his remarks that Dr. Shapiro would have told my patient with 9-mm scotopic pupils that she was at no greater risk than if her pupils were 5 mm. This is, no doubt, what the patient from St. Louis Co., Mo., was told. He claimed that he suffered "significant and permanent disturbances in his vision, particularly at night." A jury there agreed, and he was given a $3 million award. Had he simply been told his pupils put him at a potentially higher risk, he most likely would not have had a case worth pursuing. Dr. Shapiro's reading of my comments concludes that if "we dissuade large-pupil patients from undergoing laser vision correction, we may be unfairly limiting this group of patients from enjoying its benefits." The proper role of the surgeon is not to dissuade or encourage but present the information as fairly as possible. I certainly mention that some studies show pupil size is not predictive of NVC, while others show it may well be a factor. To simply inform these patients that their larger-than-normal pupils may potentially increase their risk of NVC, I feel, is in their best interest. The favorable experience with wavefront-based surgery and the ability to minimize symptoms with drops are helpful in putting this risk in perspective. The surgeon simply has to decide does he want to be more cautious or less cautious on this issue. I think it is considered standard of care to measure the pupil in dim light. Why bother measuring it if we don't discuss the possible implicationswith the patient?
Finally, Dr. William Trattler has completed an interesting study objectively measuring starburst size, and this study finds a strong correlation with pupil size and a favorable influence with wavefront-based surgery. Even with wavefront-based surgery, the patients with larger pupils reported larger starbursts and the difference was statistically significant. This paper will be submitted for publication very soon.
James J. Salz, MD
1. Schallhorn SC, Kaupp SE, Tanzer DJ, et al. Pupil size and quality of vision after LASIK. Ophthalmology 2003;110:1606-1614.
2. Netto MV, Ambrosio R, Wilson SE. Pupil Size in Refractive Surgery Candidates. J Refract Surgery 2004; 20:337-342.
3. Twa MD, Bailey MD, Hayes J, et al. Estimation of pupil size by digital photography. J Cataract Refract Surg 2004;30:381-389.
4. Pop M, Payette Y. Risk factors for night vision complaints after LASIK for myopia. Ophthalmology 2004; 111:3-10.
5. Brown SM, Khanani AM, Xu KT. Day to day variability of the dark-adapted pupil diameter. J Cataract Refract Surg 2004;30:639-644.
6. Roberts CW, Koester CJ. Optical zone diameters for photorefractive corneal surgery. Invest Ophthal & Vis Sci 1993;34:2275-2281.
7. Klonos GG, Pallikris J, Fitzke FW. A computer model for predicting image quality after photorefractive keratectomy J Refract Surg. 1996 Feb;12:S280-4.
8. Maguen E, Schlanger J, Wong J, Salz J. Relationship between total aberrations (RMS) and pupil size in ametropic eyes with the LADARWave System. Invest Ophthalmol Vis Sci 2003;44:E-Abstract 2686.
9. Maguen E, Schlanger J, Wong J, Salz J. Relationship between total aberrations (RMS) and pupil size in ametropic eyes with the LADARWave System. ASCRS Symposium on Cataract, IOL and Refractive Surgery 2003;Abstract 6.
10. Randazzo A, Nizzola F, Rossetti L, et al. Pharmocologic management of night vision disturbances after refractive surgery; results of a randomized clinical trial. J Cataract Refract Surg 2005;31:1764-1772.
11. Chalita MR, Chavala S, Xu M, et al. Wavefront analysis in post-LASIK eyes and its correlation with visual symptoms, refraction, and topography. Ophthalmology 2004;111:447-453.
12. Helgeson A, Hjortdal J, Ehlers N. Pupil size and night vision disturbances after LASIK for myopia. Acta Ophthalmol Scan 2004;82:454-460.