Approaches to cataract surgery vary in patients with previous refractive surgery

John A. Hovanesian, MD, FACS, interviews Douglas D. Koch, MD, about performing cataract surgery on patients with prior refractive surgery.

John A. Hovanesian

John A. Hovanesian, MD, FACS: Today I am speaking with Douglas D. Koch, MD, on the topic of how to approach cataract surgery in the patient with prior refractive surgery. Dr. Koch, thanks for joining us.

Douglas D. Koch, MD: Thank you for inviting me to talk on this topic, which has been an interest of mine for probably 20 years. First, let me just make a quick comment, which is that there is now online at the American Society of Cataract and Refractive Surgery Web site a calculator called the Post-Refractive Surgery IOL Calculator, available to all ophthalmologists. It was developed by Warren E. Hill, MD, Li Wang, MD, PhD, and myself. It is on the home page of the ASCRS Web site in the left column, and it has three different spreadsheets: one for the patient who has undergone myopic PRK or LASIK, one for the patient who has undergone hyperopic PRK or LASIK, and one for the patient who has undergone radial keratotomy.

Dr. Hovanesian: That is great. It is meant to be a comprehensive place so that anyone can use it. And does one have to be a member of ASCRS to use that resource?

Dr. Koch: No. This is open to all ophthalmologists worldwide.

Dr. Hovanesian: Why is it so difficult to measure the cornea of post-refractive surgery patients?

Douglas D. Koch

Dr. Koch: There are two reasons that it is difficult to obtain good corneal measurements in patients who have had prior refractive surgery. We must remember that the cornea has two curvatures, a front and a back, and when we use a standard keratometer or topographer, we are only measuring the front, but we are giving an overall value for corneal power. That overall value is based on some assumption about what the back corneal power is. If you subject a patient to PRK or LASIK, you change the front corneal power but not the back, and so you alter the relationship. These devices use a different index of refraction to compensate for the back corneal power, but if you change the relationship between front and back, that index of refraction is no longer valid.

Unfortunately, because different levels of correction change the front in different magnitudes relative to the back, there is no new single value for refractive index you can just plug into those machines to get your true corneal power. So what we are doing is measuring the front and having to make a sophisticated guess about the back, and that guess is often not as accurate as we would like it to be.

Dr. Hovanesian: And so topography really is not the ideal approach.

Dr. Koch: Topography is still valuable. It has been the mainstay of the approaches that we have used, and we have developed three different formulas that are available on the ASCRS site. But it is not the only way to go about it, and because of the fact that these corneal topographic readings can be inaccurate, corneal topographic values suffer from a second problem, which is that the uniformity of correction over the front surface of the cornea is now more variable than it was in those patients before refractive surgery. Fortunately, this problem is diminished with wavefront technology, but it still exists, so that you now have a bit of a quandary about which measurement to use that truly reflects even anterior curvature.

Dr. Hovanesian: And what about other instruments, newer instruments such as the Pentacam (Oculus)?

Dr. Koch: The Pentacam shows promise for offering another option for measuring corneal power, and it offers a posterior corneal curvature measurement that is more accurate than we have had heretofore. However, because it is an elevation device and elevation devices do not do as good a job at measuring curvature as reflective devices such as Placido technology, even the Pentacam is going to be limited in its accuracy, and I would be uncomfortable relying on it as my sole measure in these types of patients. The new Galilei device (Ziemer Ophthalmic Systems) shows great promise because it uses two cameras for its pachymetric measurements and also has a Placido to maximize the accuracy of front curvature measurements.

Dr. Hovanesian: So you make multiple measurements or take multiple approaches.

Dr. Koch: Absolutely right. You want to acquire as many data points as you can about the patient’s prior information. You try to acquire as much data as you can at the time that the patient presents for surgery, and then you plug it into all these different formulas that are now available.

Historical information

Dr. Hovanesian: Let’s talk about historical information. It is said that if you can obtain information from the time of prior refractive surgery with preoperative keratometry and then the change in refraction shortly after stabilization was achieved, you would have a good measure of what the corneal shape should be. Is that correct?

Dr. Koch: It is a reasonable approach, and we traditionally have thought that this is the most accurate approach. However, we quickly learned that this approach is actually subject to a high degree of error, and the sources of error include the fact that you are relying on K readings that are taken by a technician in somebody else’s office many years ago. We do not know if that keratometer was accurately calibrated or how good the technician was in obtaining those measurements.

Dr. Hovanesian: Exactly.

Dr. Koch: We do not sometimes know the exact amount of refractive correction that was achieved for the patient. Did the patient really have a 4 D correction, or was it 3 D or was it 5 D? On the ASCRS Web site, we have included three formulas that rely solely on historical data, but the problem with these historical formulas is that each of them depends upon the accuracy of the preoperative readings at what I call a 1:1 ratio, which means a 1 D error in historical information results in a 1 D error in your calculation. So the advantage of some of the newer approaches is that they tend to diminish the potential error that is introduced by having grossly erroneous historical information.

Dr. Hovanesian: One other potential source of error in historical information is that the cornea potentially changes over the years between the refractive surgery and the cataract surgery. How much does that happen in the real world?

Dr. Koch: That is a great point. It probably does not happen a lot in the PRK or LASIK eye, but it is a huge problem in the RK eye, and I basically have completely abandoned historical approaches when calculating IOL power for the post-RK eye.

Dr. Hovanesian: Another problem with the post-RK eye is that both the anterior and the posterior curvatures change, unlike the laser refractive patient you mention. So one theoretically has to take into account the posterior curvature differently than in the LASIK patient. Do you agree with that?

Dr. Koch: That is absolutely correct, and because we do not know how much the posterior curvature has changed — it may have changed a little in one patient and a lot in another — that introduces a major source of error and an ongoing source of frustration in accurately calculating these post-RK eyes.

Dr. Hovanesian: Let’s go through some examples of how you handle specific patients. Let’s start with the RK patient who had surgery 10 or 15 years ago who does not have historical information. The physician who did surgery does not have the records. How do you start with this patient?

Dr. Koch: I start in all of these patients by trying to obtain good corneal topography measurements. We obtain them in our own practice with three devices, the Humphrey Atlas, the EyeSys and, recently, the Galilei. There are several other devices that are also very good. But what you are looking for is a measurement that averages central corneal curvature over reasonable zones, such as 3 mm or 4 mm, to get a broader, more representative value that hopefully will truly reflect anterior corneal curvature.

In the RK eye, that is basically all you can acquire in terms of data that you can use for the calculation, other than your standard axial length measurements. You then go to your formulas. And before we go to the formulas, I want to mention a second problem that is introduced in these eyes. This problem is that the third-generation formulas, with the exception of the Haigis formula, use corneal curvature in two ways.

They use it to calculate the refractive power of the eye, but they also use it to calculate the effective lens position of the IOL. And what this means is if you take a patient who had a 45 D cornea before surgery and has had a 10 D LASIK and now has a 35 D cornea, if you use that value of 35 D in the portion of the IOL calculation formula that predicts effective lens position, it is going to say that the effective lens position is relatively anterior, and it is going to tend to significantly undercorrect the patient.

In other words, it is going to select an IOL that is too weak. So we have to compensate for this error that is introduced in the calculation of effective lens position by one of three ways. Dr. Wang and I published tables in the November 2002 edition of Journal of Cataract & Refractive Surgery that show how you can work with the Hoffer, SRKT and Holladay 1 formulas to adjust for this. You can also use the Haigis formula, which does not use K readings, in order to calculate effective lens position.

Or you can use the Holladay 2 formula, which actually allows you to enter either pre-LASIK or pre-RK K values in the formula. If these values are unknown, you can click on the “Prior RK,” etc. box, and it will default to a value of around 43.9 D for that portion of the formula that calculates effective lens position. This is a critical piece for all of these eyes that have undergone myopic refractive surgery, and the ASCRS spreadsheet effectively takes care of this problem for you so you do not have to worry about it.

Dr. Hovanesian: At what level of preoperative myopia would you really start to be worried about that effect on the measure or estimated effective lens position?

Dr. Koch: I look at it on the basis of what the corneal curvature measurements show me at the time the patient presents for surgery. If the patient has a 38-D cornea, I am worried that the effective lens position calculation could be erroneous. If the patient comes in with a 42-D cornea, then because the default for these formulas is 44 D, there is probably not going to be much of a problem.

But to finish the RK cornea, we obtain our topography readings and plug those into the Web site, and the formulas then are used. They incorporate the adjustment for effective lens position. The calculated power obviously has limitations. In selecting the IOL, I aim for a little myopia to account for potential inaccuracies, particularly if there has been a history of hyperopic progression in these patients. I also give these patients extensive warnings about the potential inaccuracies in these calculations.

Dr. Hovanesian: Have you found that the use of a contact lens over-refraction in post-RK eyes is not particularly useful?

Dr. Koch: We have been frustrated in general with the accuracy of contact lens over-refractions. But I confess that we have not taken the time to try to obtain a very good contact lens fit, which could potentially make this more accurate. There is at least one study that suggests that certain reverse aspheric designs might actually give a better reading. But we have abandoned the contact lens over-refraction approach entirely.

Dr. Hovanesian: So you do not measure it at all in your patients?

Dr. Koch: Never.

Dr. Hovanesian: Do you think that there is a medicolegal risk in not doing so, given that it is a bit of information that is theoretically available to you and that you are not collecting?

Dr. Koch: Because the literature almost uniformly indicates that contact lens over-refractions are among the least accurate approaches, I cannot imagine that there is a medicolegal risk with that.

Dr. Hovanesian: That sounds reasonable. So anything else for the post-RK eye that we should think about?

Dr. Koch: Other than surgical elements, such as where you place the RK incisions, you should make sure if you are making a clear corneal incision that there is adequate space between the RK incision so that you do not have your wound intersect an RK incision and then going back into the sclera, if necessary, if the spacing of the RK incisions does not permit a clear corneal incision to be placed between them.

Dr. Hovanesian: A very good suggestion, certainly. So on patients with prior myopic or hyperopic laser refractive surgery, do you distinguish in your approach between these two?

Dr. Koch: Yes. There does seem to be a significant difference between these two. First of all, certain formulas were validated only for myopic patients. Different topographic values seem to be useful in one and not as useful in the other. The hyperopic eye usually is easier because the corrections were smaller, and if you get a good average reading over the central part of the cornea, you tend not to have to make much of an adjustment in your corneal power, and the formulas on the Web site reflect that.

Dr. Hovanesian: It can be more challenging, I found, in hyperopes because you do not always have a uniform central power. There is some variability, so you are doing some averaging, and I worry about the accuracy of doing that. These formulas seem to work that out.

Dr. Koch: Not completely. You have raised a great point. They are not perfect in this regard, and the greater the variability over the central cornea, the more difficult it is to know what value actually reflects the “average” corneal power of that eye. The other problem with a hyperopic eye is, I think in general they are probably a little bit trickier to enhance if you have a postoperative error that the patient wants corrected, a little tougher to know how much plus or minus you put in laser, in my experience, to correct the post-hyperopic eye as opposed to the post-myopic eye.

Dr. Hovanesian: That seems very true. I agree with you. So in post- myopic excimer treatments, is there anything different?

Dr. Koch: There are a lot of formulas available, and a lot of people have done wonderful work in this area. We, unfortunately, have not been able to include every formula in the ASCRS Web site. We tried to validate the formulas that we have used based on the literature and on a large number of patients from the practices of Dr. Hill and myself. I like to think of the formulas in three categories. There are those that rely purely upon historical data, such as clinical history. Another category is those that require you to obtain measurements at the time the patient presents for surgery, but you modify those measurements a little bit based on historical information, specifically on the amount of refractive change that was induced.

Dr. Hovanesian: Which formulas fall into this category?

Dr. Koch: The Masket formula, and Dr. Hill has done a modification of that. We have a formula that uses a modification of that using both the EyeSys and the Atlas, where we take the topographic readings that are obtained and then reduce them somewhat, depending upon the amount of refractive change induced by the PRK or LASIK.

And then there is a third category, and that consists of formulas that do not use any preoperative information, that rely solely upon the data that are acquired at the time the patient presents for surgery. This includes the Haigis L and Shammas formulas and, again, a formula that Dr. Wang and I have developed.

How to approach the patient

Dr. Hovanesian: If you see a patient who has previously had refractive surgery, what sort of discussion do you have with them? Let’s assume that we are talking about a monofocal IOL.

Dr. Koch: The first thing you want to figure out is what type of refractive correction the patient desires. Presumably, because these patients have undergone refractive surgery, they want at least one eye to be corrected for distance. On the rarest of occasions that will not be true, but in general, they are going to want to have good distance vision in one eye. And then you have to make the decision about whether or not to do monovision and whether they have had monovision and how to go through that whole exercise.

It is essential to point out to them the inaccuracy of the calculations. You could even potentially print out the results from the Web site of the calculations and show the patient the variability of the lenses that are recommended by the formulas so the patient understands the extent of the uncertainty here. And then you have to go over carefully with the patients the options that will exist postoperatively, and I think it is helpful to do sufficient preoperative testing to know whether or not additional corneal refractive surgery is an option. Is the cornea thick enough? Is the tear film adequate? Are there any other problems that might preclude additional corneal refractive surgery?

So you can tell the patient that there is the likelihood that something will be needed, what the options will be and, importantly, what the cost will be to the patient so that the patient does not expect a free postoperative laser procedure to correct for the residual refractive error — unless you are offering it to them.

Dr. Hovanesian: Of the available presbyopia-correcting lenses, do you think any of them are well-suited to these types of patients?

Dr. Koch: Many surgeons have been using multifocal and accommodating IOLs in these patients. My experience is actually quite limited. I have been working primarily with monofocal designs. Because most of these patients have undergone myopic procedures, they have a lot of positive asphericity, so I like to implant a monofocal IOL that has negative asphericity. The two that are available in the U.S. market are the Tecnis (Advanced Medical Optics) and the AcrySof IQ (Alcon).

If we want to give them a broader range of correction, then monovision comes into play and can be successful using this same type of IOL in order to try to get the clearest, crispest vision for distance and near. I might also point out that it is sometimes difficult to know what to do with the astigmatism of these patients because the corneal astigmatism may or may not actually reflect what the refractive astigmatism will be. If there is any uncertainty, I tell patients, “We’re not going to correct your astigmatism at the time of surgery. We’re going to wait and evaluate that as part of the overall evaluation of the refractive accuracy so that postoperative options could include relaxing incisions if the spherical correction is sufficiently accurate.”

Dr. Hovanesian: Do I understand correctly that you do not do any type of astigmatic correction at the time of cataract surgery in patients who have had prior refractive surgery?

Dr. Koch: If the topography, particularly from two different devices, is consistent and clear in demonstrating more than 1 D of corneal astigmatism, I will go ahead and make those corrections, but if there is any uncertainty or any significant difference between the two topography devices or the topography devices and the IOLMaster (Carl Zeiss Meditec), I often will defer that pending a postoperative refraction, and then base the treatment more on the postoperative refraction than on specific corneal values.

Dr. Hovanesian: And when you have to do postoperative correction, let’s say a patient has a 2 D spherical surprise with little astigmatism, what is your method of choice, whether it is hyperopic or myopic?

Dr. Koch: For most of these patients I am doing PRK. These are older patients in general, and they do well with surface ablation. There is minimal risk. The loss of a couple of days out of their normal activities does not seem to be an excessive burden, and I tend to be reluctant to lift a LASIK flap in any case in most eyes.

Dr. Hovanesian: And certainly the older patients with flaps are more prone toward difficulties with ocular surface dryness and epithelial ingrowth, as well as other irregularities of those flaps.

Dr. Koch: I completely agree.

Dr. Hovanesian: When do you find that piggyback IOLs are a good choice?

Dr. Koch: That is a great question. Piggyback IOLs are an excellent choice in patients in whom corneal refractive surgery is not a reasonable option. Either the cornea is too thin, it is too dry or it has an irregular topography that may even be suspicious for early ectasia.

I would like to make a final comment about the calculator for the myopic and post-myopic patient. Because there are three categories of formulas, you are going to get a whole series of IOL powers: those that are predicted with clinical history, those in the middle column that are predicted using a mix of historical data and current measurements, and those that are not relying on clinical history data at all.

In our experience, the second and the third categories are more accurate. So if there is variability, we like to use formulas such as the Haigis L, the ones that we have developed with topography, the Masket, and generally those in the second and third categories to base the IOL selection. Nevertheless, the historical data can be useful. Sometimes they lead us in a direction that we might not go, but by and large, I think that because of the concern about accuracy of historical data, it is better to rely more upon measurements that are obtained at the time that the patient presents for surgery.

Dr. Hovanesian: Any final advice for approaching these patients?

Dr. Koch: They take more time. They are more demanding. They have already indicated that they want perfect vision, and it is a real source of frustration for these patients if it is not provided. You have to be sympathetic with them, even though they seem demanding, but they went into this whole process expecting high-quality vision for their lifetime. So it is our responsibility to try to give them the best correction possible, as well as a level of understanding and education that helps them go through this process as comfortably and happily as possible.

Dr. Hovanesian: Dr. Koch, thanks very much for joining us.

For more information:

• John A. Hovanesian, MD, FACS, can be reached at Harvard Eye Associates, 24401 Calle De La Louisa, Suite 300, Laguna Hills, CA 92653; 949-951-2020; fax: 949-380-7856; e-mail: drhovanesian@harvardeye.com. Dr. Hovanesian has no direct financial interest in the products mentioned in this article, nor is he a paid consultant for any companies mentioned.

• Douglas D. Koch, MD, can be reached at Cullen Eye Institute Baylor College of Medicine, 6550 Fannin, NC 205, Houston, TX 77030; 713-798-6443; e-mail: dkoch@bcm.tmc.edu. Dr. Koch is a consultant for Alcon and Advanced Medical Optics.

Reference:

• Wang L, Jackson DW, Koch DD. Methods of estimating corneal refractive power after hyperopic laser in situ keratomileusis. J Cataract Refract Surg. 2002;28(6):954-961.