A EUREQUO Moment

It’s almost eight years since Zoltán Nagy performed the first femtosecond laser-assisted capsulotomy in a human patient, ushering in the age of FLACS – femtosecond laser-assisted cataract surgery (1). Since then, the number of FLACS procedures performed has grown year on year – but the percentage of cataract surgeons adopting the laser has remained stubbornly in the single digits (2).

The principal attraction of femtosecond lasers is that they produce reproducibly round capsulotomies at exactly the size specified by the surgeon (3), something that’s believed to be crucial for achieving the best possible outcomes in patients receiving “premium” intraocular lenses (IOLs). Many surgeons talk of an element of keeping your hand in – when performing continuous curvilinear capsulorhexes (CCCs), a femtosecond laser should mean you can perform a capsulotomy that’s more perfectly circular and well-centered than “the high volume surgeon down the street” can do with his or her hands. Given that there’s a certain “Goldilocks” diameter to capsulorhexes – too small risks anterior capsule fibrosis and hyperopic IOL shifts, whereas too big risks lens tilt/decentration, posterior capsule opacification, and myopic postoperative error (4) – you can see the attraction of programming in the ’rhexis on the screen, then just clicking to have it performed for you.

The second killer feature on offer is that the laser can pre-fragment the patient’s lens, something that should make phacoemulsification easier and faster than manual phacoemulsification; also, as less ultrasound energy is required, fewer endothelial cells should be lost (4). There’s also the ability to make corneal relaxing incisions to correct astigmatism with the laser. What’s not to like?

Part of the reason why the FLACS adoption rate today is still under 10 percent is cost. A typical femtosecond laser platform can easily cost €500,000 – but beyond that, even at face value, femtosecond lasers don’t represent a total panacea for cataract surgeons. Laser capsulotomies take longer to perform than manual capsulorhexes (slowing throughput), the capsulotomy isn’t as strong, and early femtosecond lasers gained a certain reputation. Although some of the issues early femtosecond laser users experienced – like the occasional incomplete capsulotomies and the fact that anterior capsular tears seemed to occur significantly more frequently in patients who underwent FLACS (5) – are far less of a problem with the newest femtosecond laser platforms, it is going to take a while to erase those doubts. There’s also a learning curve, which is usually associated with an increased rate of surgical complications; the docking and use of the laser introduces unique risks; and femtosecond lasers are still not indicated for use in patients with small pupils (6). Although FLACS surgery is associated with less corneal endothelial cell damage than conventional cataract surgery (7), there’s some concern about the fact that the femtosecond laser still introduces energy (and heat) into the eye, which may affect intraocular structures like corneal endothelial cells and the iris (8). If we return to the “keeping your hand in” proposition above; that’s fine until a patient appears with something like small pupil – a contraindication for femtosecond laser use. So if you're “out of practice”…

In terms of anything approaching a clinical trial of man vs. machine in cataract surgery, there have been a number of evaluations (reviewed in [4]) – but all have been beset with a number of limitations: small patient numbers, no possibility of masking the procedures used, and often, a limited set of outcomes being assessed. In any event, the literature to date (albeit with the limitations listed above) has failed to demonstrate any significant differences in visual and safety outcomes between FLACS and manual CCC/ standard phaco cataract surgery (4).

“The nature of registry data is that it comes from the real world and it should reflect the reality of clinical practice.”

But there’s another way of getting robust clinical evidence of the safety and efficacy of manual vs. FLACS procedures: a registry of real-world patients and their real-world outcomes. Clinical evaluations tend to be regimented in their design and execution, with specific inclusion and exclusion criteria. The nature of registry data is that it comes from the real world and it should reflect the reality of clinical practice.

However, a registry’s usefulness is determined by the amount of information it contains – the more, the better – and that means it needs to be run under the auspices of an international organization with the membership, funds and infrastructure to do it properly. And in the case of FLACS, the organization that stepped up was the European Society of Cataract and Refractive Surgeons (ESCRS). We interviewed ESCRS past president and current board member Peter Barry about their endeavor: EUREQUO.

How did the EUREQUO platform come about?

The European Registry of Quality Outcomes in Cataract and Refractive Surgery (EUREQUO) platform was preceded by the European Cataract Outcome Study (9), which was an offshoot of the Swedish cataract registry. That system went back 25 years; you could voluntarily enter your clinic, have your cataract surgical outcomes monitored anonymously by a central office in Sweden, and have the ability to compare your results to that of other clinics or the whole database.

Ten years ago in 2006, the ESCRS succeeded in obtaining a grant from the European Union and, with matching funds provided by the ESCRS itself, decided to set up EUREQUO. Basically, we offered clinics all over Europe the opportunity to enter their data for both cataract and refractive surgery into our registry, so that they could identify outcomes and anonymously compare their results with those of the country as a whole, those of other countries, or the entire database. We have the advantage that all reporting is anonymous, so falsifying data entry is a complete waste of time – you’re only fooling yourself.

The database also gave us the opportunity to provide benchmarks – standards of outcomes and complication rates as guidelines for cataract surgery – measuring things like complication rates, precision of biometry, postoperative stigmatism, and so on. Those benchmarks were subsequently published, and they’ve proven to be very useful tools.

We now have about 1,200,000 to 1,500,000 cataracts recorded in that database. Although there aren’t nearly as many entries for refractive surgery, we’ve reached a number we can analyze, and our first report on cataract surgery outcomes in patients with previous refractive surgery has been published (10). But EUREQUO is ongoing; we get new ideas all the time, and the database is already so massive that we can often try them out right away.

What inspired you to conduct the FLACS study?

While I was the president of the ESCRS, we decided to use the EUREQUO database for a comparison study of FLACS vs. traditional phaco. For the last five years, it has been impossible to go to any meeting, read any journal, or pick up any magazine related to cataract surgery without being inundated with information on femto. But the difficulty is that the numbers are usually quite small and investigators frequently have a vested interest in the product, so we felt there was always a subliminal message that femto was automatically better because it was so expensive and so high-tech.

We knew that the French were planning a pretty major clinical trial on femto (11), and we hope that those results will be available sometime this year. But we wanted to do a preliminary assessment to see if, in its current state of relative infancy, femto outperforms phaco. We have standard questions in the EUREQUO database, so we decided to add specific fields for femto cases and then, as precisely as possible, perform case–control study matching between prospectively entered femto patients and traditional phaco patients whose cataract surgery was entered into the database in the 2014 calendar year.

How did you conduct the study?

We recruited volunteers, and the ophthalmologists who responded were leading cataract and femto surgeons from high quality clinics and university departments around Europe. In the end, we had 16 clinics participating in 10 different countries. We insisted that each participating surgeon had completed a learning curve of at least 50 cases, but most of them had done much more than that. The patients themselves were selected by the surgeons, but we did insist that consecutive cases be submitted – otherwise, you’d simply finish up with a best-case scenario that wouldn’t have any validity whatsoever.

Our first patient was recruited on the first of December, 2013 and the last one was recruited on 31 May 2015. In total, we recruited 3,379 femto patients – and then we needed to match them. The process was very precise; patients had to be within two years of age, have exactly the same preoperative visual acuity, and score exactly the same number of yes/no boxes in the fields measuring preoperative ocular comorbidity and anticipated surgical difficulty. We ended up with 29,761 suitable cataract surgery patients – from which we matched 2,814 of the femto patients. The remaining 565 couldn’t be matched primarily because of their age and visual acuity; they were just too young, or their vision was too good.

And what did you find?

When we looked at the scoring for preoperative anticipated surgical difficulty – which, in the EUREQUO, includes previous corneal refractive surgery, white cataracts, pseudoexfoliation, previous vitrectomy, corneal opacities, and small pupils – the results were strikingly similar between the femto and the phaco patients, with one significant difference: in the femto group, 4 percent had undergone previous corneal refractive surgery, whereas in the phaco group, that number was infinitesimally small at 0.1 percent. But across all of the other fields, the anticipated surgical difficulty was essentially the same. Ocular comorbidity – glaucoma, age-related macular degeneration, diabetic retinopathy, amblyopia, and “other” – was also remarkably similar. So the two groups were definitely well-matched prior to surgery.

In terms of perioperative complications, the femto group initially appeared to do a little worse – but that was simply because there were extra fields (like difficulties with docking or minor capsular problems) specific to the procedure. When you excluded those, femto did slightly better in terms of the overall perioperative surgical complication rate, but the specific complications were virtually the same in the two groups.

But when it came to postoperative complications, femto did slightly worse overall. In particular, of the 10 specific fields in the EUREQUO database, femto performed meaningfully worse with respect to corneal edema, early posterior capsule opacification (PCO), and uveitis requiring treatment. Those results are not just statistically, but also clinically significant.

What about the patients’ visual outcomes?

When you look at the postoperative corrected distance visual acuity, the phaco patients did a little better overall. The differences were small; they just did slightly better in every visual acuity category. More importantly, we looked at what percentage of patients had improved visual acuity compared to their preoperative status, what percentage had no change, and what percentage had worse visual acuity after the surgery than they had before it. In those categories, the femto group did slightly worse in each case – one striking figure was that 3.3 percent of femto patients had worse visual acuity after the cataract surgery than before, compared to 1.3 percent of the phaco group. Your immediate response might be, “Of course the femto group is going to be worse than the phaco group, because you’ve already told us that a greater number of them had undergone previous corneal or refractive surgery.” But that actually isn’t the reason. When we looked into it, the worse vision was due to postoperative complications, not the previous corneal refractive surgery, and the specific complications at fault were corneal edema, early PCO and uveitis.

We were also concerned with cylinder. When we measured patients with a postoperative cylinder of 1.5 D or greater, femto did meaningfully better – twice as well at 9 percent when compared with the phaco group at 18.5 percent. And it wasn’t because of the more frequent use of premium IOLs in the femto group. After subtracting all of the patients who had received toric IOLs, undergone concurrent refractive surgery, or had a history of previous corneal refractive surgery, the differences were precisely the same – 9.2 percent of the femto vs. 18 percent of the phaco group. So femto unequivocally did better in terms of postoperative astigmatism. We also used in vivo analysis to look at surgically induced astigmatism, and on all fronts, the femto group had less surgically induced astigmatism than the phaco group.

Does this mean that phaco is better on all fronts bar astigmatism?

I think that’s true, but we have to remember the limitations of what was measured. We could only compare measurements recorded by EUREQUO – and there’s nothing easier than criticizing a clinical trial for failing to answer questions it didn’t ask. We measured a laundry list of parameters, but there are still plenty we omitted – not because we considered them unimportant, but simply because there were no data in the EUREQUO interface with which to compare them. And we have to be very fair and honest here, because enthusiastic proponents of femto will identify that all of those issues are major reasons to support the technology. I think it’s critically important to represent the study’s limitations; we’re not femto-bashing, we’re just trying to say to industry, “We know you’re enthusiastic, and we know there are very enthusiastic surgeons out there, but we also know that it’s very expensive. The onus is on you to demonstrate to us that the cost is justified in terms of better outcomes.” And at its current stage of development, I would say that femto does not outperform phaco.

Do you anticipate adding any benchmarks or conditions to the database in the future?

As I said, we lost 585 patients in the FLACS study because we couldn’t match them. Now we’re looking at the approximately 3,500 femto patients in the database and subjecting them to detailed analysis – who are they? How many of them have had previous refractive surgery, and at what age? What was their vision like prior to cataract surgery? How many of them have concurrent refractive components to their cataract surgeries? How many of them have subsequent top-ups? How many achieve “super vision?” We want to know if they’re a completely different population of patients, and to understand that, we need to look at all of the information.

When we presented our data, we received a question about the relative use of premium IOLs in the two groups. The case–control study didn’t use type of lens (i.e., monofocal vs. premium) as a comparator and the audience wanted to know why. The answer to that question is twofold: one, that we would dearly have liked to match the patients that way, but we couldn’t, because only about 1 percent of database entries had received a premium IOL; and two, that in this study, only 55 percent of the femto patients received monofocal IOLs, compared with over 99 percent of the phaco group. We’re now in the process of comparing those two groups as fast as possible, which was something we simply didn’t think of earlier because we were sidetracked by the fact that we couldn’t include IOL type as a matching factor. When we publish the peer-reviewed paper, it will include the answer to that question.

What kind of patient would be an ideal candidate for FLACS?

Based on the study, I don’t think you can identify the ideal femto patient. But my colleague Jose Güell presented on femto in difficult cases at the 2015 ESCRS meeting in Barcelona (12) – addressing issues like pseudoexfoliation, partially subluxated lenses, or “difficult” cataracts. The femto procedure minimizes perioperative manipulation, which, in his eyes, makes it a safer procedure in that subgroup. But other than that, I think doctors shouldn’t feel that they as surgeons, their clinics, or their institutions, are in some way inferior because they are currently unable to offer femtosecond laser-assisted surgery. And they should communicate to their patients that, even if a technique is more “high-tech,” they shouldn’t assume it’s automatically better.