What effect does CXL have after 10 years? This August, ELZA doctor and researcher, Emilio Torres-Netto MD was interviewed by TouchOphthalmology on a recent congress presentation he gave on 10-year cross-linking follow-up data collated from two treatment centers in Zurich, and was asked about what the research group found.
What is the rationale for corneal cross-linking in progressive keratoconus?
Before your study, how much long-term data on the use of corneal cross-linking were available?
Not that much actually, very limited information was available for us on long-term follow-up. There are now more than 500 papers in PubMed about cross-linking since it was established in 2002/3… but in the long-term, actually, there isn’t that much. We do have the Sienna study with quite a long follow-up and also we have a Dresden follow-up with 10 years of data, with similar results. The overall result is quite similar to ours, so they had in Dresden their Kmax (i.e. the maximum keratometry) reducing by around 0.1 logMAR; more or less was what we found, and the main difference is that our series shows a bit more stability of refractive outcomes, so both in terms of spherical equivalent and also astigmatism. But our findings from patients in Zurich and the findings from patients in Dresden (published previously) agree a lot between them. We decided to evaluate this data because of course, it can bring a greater understanding: more information about what can happen in the long term, and a better understanding of how we could potentially manage the patients better.
The data that we had were collected from surgeries between 2004 and 2019. The surgeries were performed at IROC and the ELZA institute, both in Switzerland. We had more than 400 surgeries, but to evaluate the long term, after 10 years, you can imagine we could not reach all of those patients again, some of them did not have the follow-up completed so we ended up analyzing 50 eyes! And what we saw, was that, as expected, cross-linking could stabilize keratoconus progression. We had around 50 of the patients stable visual acuity, and this was more or less one-line stability and 36 percent actually had improvement of more than one line after 10 years. This really surprised us, because that was not initially the idea to improve vision, the idea is to stabilize it. But we could see that many patients had also an improvement in visual acuity from baseline levels. So 10 years after cross thinking, we saw a decrease in keratometry (the Kmax) of around 1.5 diopters in 10 years. That’s quite a lot considering that there is a long time interval but I would say what we were most interested in was understanding how visual acuity in the long term would change. So this is a challenge. We have to answer patients who ask: “Can my visual acuity can stay over a long period?”. We see that 50 percent stay stable. But we would like to see if there was anything that we could have predicted, such as worsening of visual acuity for example. We looked for any risk factor of changing visual acuity, but we could not find any relationship between VA loss and age or also with any Scheimpflug measurements like game maximum keratometry or any other parameters… but we did see some correlations with pre-operative visual acuity. So in other words, the most severe visual worsening that we saw was in those eyes with the best pre-operative visual acuity. There are other studies saying this as well; this confirms that good pre-operative visual activity can be a risk factor for visual acuity changes so that was maybe what we find most interesting.
What adverse effects were seen and how many repeat procedures were needed?
Of course, we had some failures of the treatment and we could not evaluate all patients treated thanks to loss to follow-up, but we did evaluate treatment failure which is a long-term complication. So from 42 eyes that had not had other refractive procedures, we saw that two eyes required re-cross-linking (so a second procedure). One of them was 32 and the other 19 years old, and then another eye needed a keratoplasty (DALK) and that patient was also young at the time of course (I think he was 25 years old). Another three eyes progress (an increase in Kmax) in the overall period, so we could say that, overall, cross-linking had stopped the progression of keratoconus in around 85 percent of patients after 10 years. That was our main finding.
What will be the impact of these findings on clinical practice?
Once we understand better the outcomes of CXL after 10 years, we are better able to improve our treatments. One of the curiosities we found was that the two youngest patients in our case series (both of them were nine years old on the day of the surgery). They remained stable after 10 years of cross-linking. So such small observations are also very valid because there is a discussion about whether the collagen turnover from the cornea means that the cross-linking effect might be lost over the long term, and this might mean that younger age has a negative impact on post-cross-linking outcomes. But observing that two 9-year-old children (at the time of treatment) were still stable 10 years after cross-linking, gives us some very important insight on the long-term stability of corneal cross-linking.