Retinal detachment in myopic RLE overstated

Still regularly overstated, here is why re-evaluating the risk of refractive lens exchange is so important.

For 20 years, the topic of refractive lens exchange has remained highly controversial. It has recently been written about by myself and others as an idea whose time has finally come, but many still feel strongly that it must be avoided. The polarizing issue is the retinal detachment risk, which seems increasingly to be attributed to the surgical intervention alone. This is a critical oversight with implications that reach far beyond the refractive lens exchange debate.

RLE: not for myopia only

There are fewer objections to refractive lens exchange (RLE) now than in the early 1980s when Verzella first proposed the concept of lens extraction for correction of high myopia early in the IOL era. Back then, the idea was almost unanimously rejected. Even those of us who embraced the concept of RLE saw it as a good idea ahead of its time. Aware that we did not yet have the techniques to consistently deliver the corrective potential of this approach, we turned our attention to developing and refining those techniques in refractive cataract surgery.

It was unfortunate that RLE was initially proposed as a treatment exclusively for high myopia. Critics at the time, rightly in my view, pointed to the unacceptably high risks of retinal detachment (RD) in this group as the primary reason for rejecting the procedure. Since then much has changed. Our surgical techniques are far superior, with significantly fewer complications. But the most important change is that those for whom RLE was initially proposed — prepresbyopic high myopes — are for the most part no longer considered candidates by virtue of better alternatives. RLE is emerging as a procedure primarily for presbyopic hyperopes.

Nevertheless, RD remains closely linked with RLE in the minds of many. Much of this stigma is based upon genuine and appropriate concern. However, when the risk of RD is used as a sweeping indictment of RLE in general – not just in high myopes – then something is amiss.

Component risk assessment

In recent months I have been challenged several times by colleagues on the idea of RLE. Most often, they refer directly to the report by Joseph Colin, MD, of his 7-year follow-up on a prospective study of RLE in high myopes. Referring to one datum in particular, they warn that the risk of RD with RLE is 8.1% and rising dramatically. These warnings – based on the statistical sound bite of 8.1% risk of RD in high myopes – are not explicitly reserved for high myopes. This headline statistic has almost taken on a life of its own. Yet it rests upon a faulty interpretation of Dr. Colin’s numbers, in my view. My own reading of this data led me to decidedly different conclusions as to just what that 8.1% signifies. Here is why: Total risk = pathogenic risk + iatrogenic risk.

Retinal detachment is one of those morbidities with two potential risk factors: pathogenic risks and iatrogenic risks.

The total risk of RD is the sum of these two component risk factors. The pathogenic risk for the cohort under study is determined from epidemiological data. Colin used data from Goldberg’s 1989 appraisal of this issue, who had turned to Perkins’ 1979 study. According to Perkins, the underlying risk of RD is 0.68% per year for myopia of greater than –10 D. (Colin applied this value to his cohort of greater than –12 D.) The other risk categories for RD used for this report are worth noting:

————————————————————
Myopia

-10 and greater = 0.68% risk/year*
-5 D to -9.75 D = 0.075% risk/year*
0 D to -4.75 D = 0.015% risk/year*

* Note the enormous increased risk in myopia > -10 D when compared with low/moderate myopia.

Hyperopia

0 D to + 4.75 D = 0.002% risk/year
>+5.00 D = 0.0015% risk/year**

** Hyperopia seems to correlate with progressively decreased risk of RD at least up to +5D.

————————————————————

The iatrogenic risk component, with a default value of zero, is activated by intervention. Thereafter, both components must each be assigned a value in this equation. The iatrogenic risk factor is determined as the difference between the total risk and its pathogenic component.

Applying this component risk assessment to Colin’s study, the following results are evident through a 7-year follow-up period: There was an 8.1% total incidence of RD (four cases in three patients over 7 years). This can be divided into a 4.76% pathogenic risk factor for RD (0.68% per year over 7 years) plus a 3.64% iatrogenic risk factor. In previous publications, Colin had reported the 2-year (0% RD) and 4-year (1.9%, 1 case) follow-up.

More recently, 10-year follow-up data from Colin (still unpublished) reveals a 10% total incidence of RD. At first glance, this seems to reinforce the warnings. However a more measured assessment gives us the following through a 10-year follow-up period: a 10% total incidence (five cases, four patients in 10 years); a 6.8% pathogenic risk factor (0.68% per year over 10 years); and a 3.2% iatrogenic risk factor. When our goal is to assess the relative safety of a given surgical procedure, it is the iatrogenic risk factor — the additional risk specifically attributable to surgical intervention — that we are after. That risk is not 10% or 8.1% but 3.2% over a 10-year period when evaluating these data. (See Table 1)

Table 1. Component risk assessment using Perkins’ pathogenic risk data for high myopia cohort (10 D>)

*Note that the total incidence of RD is increasing with time. This is inevitable and directly related to the pathogenic risk component which is a function of time. After being activated by surgical intervention, the iatrogenic risk component may be starting its decline back toward zero as would be expected. Source: Maloney WF

Statute of limitations

Clearly the total incidence of RD in high myopes will increase with time. But this inevitably steady rise is a direct result of the pathogenic risk component, which is necessarily a function of time (0.68% risk per year).

Once activated by an intervention, the iatrogenic risk factor, which is no longer zero, adds to the underlying pathogenic risk factor and bumps up the total incidence but only for some unspecified period. As time passes after the intervention, the iatrogenic risk should wane and the underlying pathogenic risk gradually reassumes dominance. In other words, there is, at least theoretically, some statistical statute of limitations to surgically induced risk. This is what Colin’s data may be starting to show as the iatrogenic component has recently decreased from 3.64% to 3.2%.

It should be clear from Table 1 that the total incidence of RD will always necessarily overstate the iatrogenic risks by the amount of the pathogenic risk component. We can debate the accuracy of the pathogenic risk factor used — a valid debate in my opinion — but we cannot overlook it and claim to be on solid statistical ground. This is the core error inherent in the 8.1% headline figure (now 10%) that continues to bounce around.

None of us would argue that the iatrogenic risk factor in this surgery could be zero. We immediately understand that proposition to be false. Yet it is actually a greater statistical error to assigning a zero value to the pathogenic risk factor in high myopia, a retinal disease unquestionably associated with significantly higher rates of RD.

Let me stress that I do not mean to suggest that 3.2% might represent an acceptable iatrogenic risk factor in this equation. I, personally, am far from that position and have no desire to perform RLE or any other refractive surgery in high myopes. I am simply trying to give voice to what seems to me should be obvious but clearly hasn’t been: Our risk assessment is faulty when we overlook the pathogenic component.

Why it matters

This critical oversight has implications that reach far beyond this particular debate. As we continue to develop and implement other surgical corrections for high myopia — phakic IOLs in particular come to mind — this becomes a medicolegal issue with significant ramifications. Getting this right matters to all of us, not only those concerned about RLE.

At our course at the recent ASCRS meeting in San Diego, Richard Lindstrom reflected the prevailing point of view when he mentioned that he wasn’t sure that the underlying risk really mattered clinically.

“Our own study on this showed a 5% total incidence of RD and, whether my RLE procedure or the underlying myopia caused it, all of those patients still blamed me,” he emphasized.

And that is precisely the larger point. Somehow we have gotten this issue backwards. Proper accounting of the pathogenic risks – not only the iatrogenic risks – lies at the heart of informed consent. This holds true whether we perform RLE, LASIK, cataract surgery, a phakic IOL procedure, a YAG capsulotomy or even an indirect exam with scleral depression on a high myope. (Yes, I know of a lawsuit brought because a myope developed an RD following an indirect exam.) If we overlook the pathogenic risk factor, as well as that statistical statute of limitations on the iatrogenic risk factor, we increase considerably our own medicolegal risk. We inadvertently steer the patient directly toward the natural default conclusion: that all morbidity occurring postoperatively must have been surgically induced, which is a natural conclusion, but way off the mark.

Armed with an understanding of component risk, Dr. Lindstrom’s staff could have explained to highly myopic candidates that their risk of RD is a combination of pathogenic and iatrogenic risk factors. Using Perkins’ epidemiological data, for example, the patients could have been advised that the underlying risk of RD due to the retinal disease of myopia (6.8% over 10 years) was more than twice that of the additional risk posed by the RLE surgery (3.2% over 10 years). Further, by stressing the theoretical statute of limitations, the patients would learn that this difference in relative risks will steadily widen as time passes. Surely this information would have made it far less likely that those patients would necessarily blame the surgery and, by extension, the surgeon.

Perkins’ report, published in 1979, is a jarring testament to the pathogenic morbidity of myopia. His route to the 0.68% risk factor for RD is tortuous, however, and axial length would be a far better way to assess myopia than refractive error. We can debate its accuracy, but we should not continue to utterly ignore the pathogenic risk factor as we discuss surgical options and alternatives with patients. Component risks should be an integral part of the every informed consent, especially in myopic surgical candidates.

Why this statistical stumble?

How is it that the pathogenic risk component somehow faded to the background like this? It seems likely to me that it stems from our response to the initial proposal that RLE might be an appropriate treatment for high myopia. Those opposed mounted their challenge by highlighting the pathogenic risks of these eyes. Initially the argument was that these eyes have a great deal going against them already, so we should not add to that with the considerable additional risks associated with intraocular surgery. The primary emphasis initially, however, was on the underlying pathogenic risk of myopia.

Gradually, as techniques improved, the added risk of surgery might be less weighty evidence for the prosecution, but to many, the core argument still held. These eyes are inherently too risky for corrective surgery… period. It seems to me that the pathogenic risk factor faded from the equation and the total risk was more or less implicitly equated with iatrogenic risk alone. A good intention seems to have led us to a precarious medicolegal position.

Keeping our perspective

Correction of high myopia remains the Holy Grail of refractive surgery. But why? It seems to me that we are preoccupied with a demographically high-risk/low-reward subset of the population. And they have increasingly attractive alternatives.

Continuous wear contact lenses are one option. Granted, a certain percentage will be contact lens intolerant, but can we honestly tell ourselves that this option has been exhausted on behalf of this group of patients? After that, the phakic IOL seems to me a more suitable option than RLE. But let’s not forget that our patients really do depend upon us to guide them through this decision, and sometimes the best thing we can do is to strongly recommend against any surgery. These extremely myopic eyes really do have a lot going against them from the outset, and in some cases even the appropriately assessed iatrogenic risk factor is too much.

However, by shifting emphasis to the hyperopic presbyope – by far the largest segment of the adult population – we sidestep the RD controversy and focus our energies on those patients emerging as excellent candidates for RLE. Their low underlying pathogenic risk component for RD (0.002% for hyperopia > +4.75 D according to Perkins’ report) dictates the rarity of this complication here. Hyperopia apparently protects against RD, just as myopia enables it. Pragmatism may be the answer for each side of this controversy.

To conclude

As I have written here before, I see RLE as a manifestation of phase two of the IOL revolution begun by Sir Harold Ridley. The paradigm that directed us through phase one was to develop and deliver the considerable corrective capabilities of the IOL to cataract patients. It culminated in today’s refractive cataract surgery. Phase two, now just getting under way, will be directed by the principle that the corrective capability of the IOL is too compelling to remain restricted to cataract patients only.

From this perspective, RLE represents one of the natural next steps. It is a broader application of fully developed techniques. I have also stressed however, that this one small step across the 20/20 threshold can be a giant leap for even the most experienced cataract surgeon and should not be taken for granted.

I should also again stress that the rapid growth of RLE is being driven primarily by its appeal to presbyopes, most of whom are mild to moderate hyperopes.

This is my take on our present position. Before phase two is over (remember that phase one took 50 years), the IOL will be a mainstay of vision correction surgery. Spectacles will be unnecessary for most presbyopes. However, contrary to the headlines, I doubt the vast majority of baby boomers will elect surgical correction of their presbyopia, extreme makeover notwithstanding. I expect a return to reality soon.

Next month:

Let the patient choose the compromise: Custom correction for presbyopia.

References:

• Colin J, Robinet A. Clear lensectomy and implantation of low-power posterior chamber intraocular lens for the correction of high myopia. Ophthalmology. 1994;101:107-12.
• Colin J, Robinet A. Clear lensectomy and implantation of low-power posterior chamber intraocular lens for the correction of high myopia. A four-year follow up. Ophthalmology. 1997;104:73-8.
• Colin J, Robinet A, Cochoner B. Retinal Detachment after clear lens extraction for high myopia: seven-year follow-up. Ophthalmology. 1999;106:12:2281-85. discussion: 2285.
• Goldberg MF. Clear lens extraction for axial myopia. An appraisal. Ophthalmology.1987;94:5:571-82.
• Perkins E.S., Morbidity from myopia. Sight Saving Rev. 1979;49:11-19.
• Preetha R, Goel P, Patel N, et al. Clear lens extraction with intraocular lens implantation for hyperopia. J Cataract Refract Surg. 2003;29:89:5-9.
• Verzella F. Atlas of extracapsular technique, phacoemulsification and intraocular lenses. 1981: 98-107.
• Verzella F. High myopia: microsurgical extracapsular extraction of the lens for optical purposes. Presented to the Keratorefractive Society, Chicago, October 29-30, 1983.
• Verzella F. High myopia: in-the-bag refractive implantation. Ophthalmic Forum. 1985;3:3:174-5.

• Note: Thanks to Mary Archer and the faculty of our OSN Editorial Board Course “Refractive Lens Exchange – A Practical Guide” at the ASCRS meeting in San Diego. Dick Lindstrom, Skip Nichamin, Bruce Wallace and I hope to do another course at next year’s ASCRS in Washington, D.C. Apologies to those of you who did not get the handout material at the course. We had more than 200 handouts prepared, but they weren’t enough for about 300 attendees. Those who requested materials will receive them soon.