What We Should Really Tell LASIK Patients

Jack Holladay, MD, MSEE, FACS Houston

Small pupils or not, patients will lose some contrast sensitivity.

All surgeons know that, done properly, LASIK and PRK can markedly improve uncorrected Snellen acuity. But there is a flip side to these procedures of which surgeons and their patients are completely unaware. Because excimer laser manufacturers were unaware of the true physiologic shape of the human cornea, both procedures reduce contrast sensitivity in low-light conditions. In this article, I'll explain the problem, how it impacts patients, and what you can do to minimize the problems associated with it. I'll also discuss what excimer companies are working on to reduce, or eliminate, these concerns.

Excimers assume all corneas look like this, a sphere. Unfortunately, that's not the case.

Function follows form
To understand the problem, it's important to understand the impact of the cornea's shape on the light that passes through it.

The natural shape of most corneas is what's called prolate. A prolate surface is shaped something like the head of a bullet. It is steep curvature in the center, but flatter toward the periphery (see figure 1). As the cornea flattens, its dioptric power decreases. The net effect is that central and peripheral light rays all focus at a single point inside the eye, regardless of pupil size.

In a small percentage of the population, the cornea is spherical. In these patients, the focus of rays shifts depending on pupil size. In the daytime, when the pupil is 3 mm or smaller, most of the rays focus at one point. At night, when the pupil is dilated, the peripheral rays focus at a point anterior to that of the central cornea (see figure 2). This results in night myopia; the optical property is called spherical aberration.

The problem with all excimer lasers on the market today is twofold. First, the engineers assumed that the cornea is spherical rather than prolate. Second, they assumed that their job was to reshape a relatively steep sphere into a relatively flat sphere, rather than to reshape a steep prolate into a flatter prolate.

As a result, excimer lasers actually reshape prolate corneas into what is known as an oblate. An oblate is shaped like the cross-section of a hamburger bun, flatter in the center but steeper in the periphery (see figure 3). This shape is actually optically worse than a sphere, because now the peripheral rays are bent even more powerfully than in the periphery of a sphere, causing even more pronounced spherical aberration when the pupil dilates.

Unlike the haloes that result from a too-small optical zone, this problem affects every patient who undergoes an excimer laser procedure to some extent. Like the halo problem, the oblate dilemma is worst for patients whose pupils dilate widely at night.

Figure 1 (right). The prolate cornea. In this, the cornea's natural shape, light rays come to a single point of focus.

Figure 2 (left). The spherical cornea. Under dilation, peripheral light rays cause spherical aberration.

Oblation in practice
We recently tested this concept on seven patients before and after they underwent bilateral LASIK. We first tested the patients at 98 percent contrast, which is the level of the letters on a Snellen chart. Then, we decreased the contrast to 13 percent, which is roughly equivalent to reading a faded newspaper. Finally, we tested each person's contrast threshold. The contrast threshold test involves showing the patient a 20/200 size letter, then decreasing the contrast to the lowest possible level at which he can still read it. To determine the extent of the pupil's role in any qualitative vision problems, we performed these tests under conditions of high and medium brightness, and in the dark (pupil size averaged 6 mm).

You can best visualize the concepts of oblate and prolate with an ellipse. The top and bottom of the ellipse above and to the left is oblate; the sides are prolate.

On the first day postoperatively, though high contrast acuity stayed relatively stable in medium and high brightness levels, it decreased a full line in the dark. Thirteen percent contrast acuity dropped 1.4 lines in dark conditions.

At one and six months, high-contrast acuity in medium and high brightness was back to normal, and even slightly exceeded baseline levels. This is most likely due to the reduction in minification from the thicker high minus spectacles the patients wore preoperatively. In darkness, however, high-contrast acuity was still slightly below baseline, by approximately a quarter of a line, at six months. At the lower contrast level, 13 percent, acuity was even worse, almost an entire line below preoperative levels. Also, at the six-month mark, the average contrast threshold in dark conditions was 2.3 lines poorer than at baseline, though it returned to baseline under medium and high brightness levels.

It's important to remember that these patients' pupils did not dilate outside the ablation zone. Patients with pupils that do dilate this much could be expected to have even lower contrast sensitivity.

The present
Right now, practicing surgeons can do little to alter the ablation pattern of their excimer lasers. The manufacturers will have to address that in the days and months to come. Right now, our one, best recourse is good, effective patient screening and counseling.

The first step is to make sure you take an accurate measurement of pupil size. As I mentioned, the larger the pupil, the more optical aberration patients will experience from the oblate cornea effect. Large-pupil patients are also much more likely to see haloes at night, because their pupil exceeds the ablation zone.

To measure the pupil, I simply darken the room and then aim a penlight covered by a cobalt blue filter from the temporal side of the eye towards the nose. Blue light is also very inefficient at causing pupillary reflex. I then measure the pupil with a pupil gauge manufactured by Asico. I have recently begun to use the Colvard infrared Pupillometer. If the patient appears to dilate widely under dim lighting, I use this pupillometer, which is much more accurate and requires no visible light.

If the patient is a large scotopic pupil, I emphasize that the quality of vision during night driving will be poorer and that supplementary spectacles may be necessary. I also inform the patient of the possibility of haloes around lights at night. Patients who wear contact lenses often are already familiar with haloes because the optical zones of their contact lenses are also too small for their pupils. If the patient is not bothered by these haloes, I tell them the haloes from LASIK probably will be very similar and should not bother them either. If the patient seems concerned about the haloes and states he or she has never seen them, then more discussion is required.

Remember that high correction and astigmatic correction both can effectively reduce the size of the optical zone. This will intensify the halo effect.

The oblate cornea (above). The powers in the periphery become stronger sooner (bottom left map), causing spherical aberration.

A prolate cornea (right). The lack of stronger powers in the periphery (bottom left map) means no spherical aberration.

The future
Since the oblate cornea is a creation of the laser itself, only the laser companies can ultimately do away with it. Here are some of the developments underway:

New beam profiles. Companies are already working on revising their beam profiles to leave the cornea in a prolate shape. This change should be relatively simple for both broad-beam and scanning lasers to effect. Unfortunately, the new pattern will require the removal of 20 to 25 percent more tissue. This will lower the maximum amount of refractive error that surgeons can correct with a flatter and prolate cornea. For example, if -12 D is approximately the limit now, it will probably drop to -8 D if the surgeon wants an anatomical prolate cornea.

Adjustable parameters. I believe new software developments will allow surgeons to enter all the vital patient characteristics into a laser's computer--pupil size, corneal thickness, corneal asphericity and intended correction--and then optimize the correction. For example, if someone has an error slightly above the theoretical maximum of -8 D, the laser may "know" to make the cornea slightly less prolate, or to use a smaller optical zone, depending on the patient's pupil size.

Though contrast sensitivity loss is undesirable, especially in a procedure designed to improve vision, the effects are not debilitating, and many patients are very willing to make the trade-off. We have a responsibility to inform patients of this potential problem, however. We also have a responsibility to continue to push laser manufacturers to improve their software so that we can reduce these adverse effects.

Dr. Holladay, who is in private practice with the Houston Eye Associates and is the McNeese Professor of Ophthalmology at the University of Texas Medical School in Houston.