A Clear-Eyed Look at RLE

Refractive lens exchange (RLE) is becoming an increasingly common procedure. Typically, people in their early 40s and 50s, who have disposable funds, elect for the operation to address deteriorating vision or gain spectacle independence. In the short term, they appear happy with the procedure; RLE seems to gives them what they want (1). But is patient satisfaction with early postoperative outcomes lulling us into a false sense of security? Some ophthalmologists are starting to question whether RLE is as safe – and effective – as we think. Indeed, in my practice, I now try to talk patients out of it. Why? Because I believe that, for the typical RLE patient, the procedure risks adverse outcomes for little or no long-term benefit. Let’s look at the evidence.

It’s useful to consider what visual problems RLE patients seek to resolve, and to what extent current IOLs actually address those problems. Briefly, patients request RLE because they have one or more symptoms of dysfunctional lens syndrome (DLS) – presbyopia, disability glare, decreased contrast sensitivity and decreased night-time vision. DLS symptoms such as disability glare are caused by forward light scatter from an early cataract; it’s been said that if the ocular scatter index (OSI) value is greater than one, then you should consider surgical intervention (2). The assumption is that providing an IOL will improve their DLS symptoms – but I believe this deserves to be examined more closely.

My first concern is that the majority of current IOLs actually produce forward light scatter (Table 1) – so they may end up causing the very problems patients hope to escape with RLE. Further, the level of forward scatter associated with IOLs is not trivial – it is similar to that of a mild cataract. The effect has been reported by different groups who looked at various IOLs and at different follow-up times (3)(4)(5)(6); in all cases, post-RLE eyes had OSI values in excess of 1.0 – so patients who start with an OSI above 1.0 will again have an OSI above 1.0 after the RLE procedure. Results from our own studies and from the investigations of others (7)(8) have indicated that OSI can be as high as 2.3 in post-RLE eyes (Table 2, a, b). 

My second concern with RLE is that current IOLs do not restore accommodation or effectively treat presbyopia. When considering vision across distance, intermediate and near, the only systems that work are micro-monovision or multifocal IOLs. And although multifocal IOLs offer the best outcome, they come with a key disadvantage: photic phenomena. And the problem is bigger than most people realize.

Photic phenomena are among the commonest reasons for patient dissatisfaction with IOLs (9)(10)(11), but this fact is sometimes obscured by the very low self-reported incidence of dysphotopsia (0.2–1.5 percent). When you proactively ask patients about dysphotopsia symptoms, you find that the real incidence is between 20–77 percent. In other words, about half of IOL recipients have some photic phenomena (12)(13)(14)(15)(16)(17). This frequency of visual disturbance is even supported by statements on multifocal IOL package inserts, with glare and halo incidence alone reported to be about 20 percent (18)(19). 

Collectively, the weight of evidence suggests that RLE fails to improve symptoms of DLS.  So why are we recommending a procedure that has little benefit? I think the actual situation is worse still: RLE may actively increase risks for patients.

Increasing risks

We should remember that IOL implantation is not risk-free. Retinal detachment (RD) may be a particular concern. In the general population – those who have not received ocular surgery – the RD risk is very low (0.0061–0.0179 percent per year). For those who have undergone phacoemulsification the risk rises, reaching 3.6 percent over a follow-up period of four months to two years. Even more tellingly, data from over 200,000 patient records have indicated that operated eyes have a five-fold higher risk of RD following surgery compared with non-operated eyes, and this risk remains elevated for 10 years following surgery (20). And younger, more myopic patients – that is, the ones who are most likely to elect for RLE procedures – actually have a 25-fold higher risk of RD than the standard cataract patient (21). Moreover, because they are younger patients, this risk is elevated for a much longer proportion of their lifetime than an older patient.  

We also need to consider IOL decentration – which can occur in around 25 percent of all cataract patients – and dislocation. The frequency of dislocation increases over time: at 10 years there is a one percent risk of IOL dislocation requiring surgery, a 0.7 percent risk of pronounced pseudophakodonesis and a 1.4 percent risk of moderate pseudophakodonesis (22). Concerningly, the frequency of IOL dislocations seems to be increasing (23): in a retrospective cohort study of 140 eyes, the cumulative risk of late dislocation was found to be significantly higher (p<0.001) in patients who were operated on between 2002 and 2012 compared with patients who had IOLs implanted between 1992 and 2001. We must conclude that some aspect of more recent IOL procedures increases the subsequent dislocation risk; we don’t know the cause, but we should be aware of the effect – and advise our patients accordingly.

Optical quality

Finally, we should remember that the optical quality of an implanted IOL is not constant, but degrades after implantation. Acrylic IOLs (such as AcrySof, which has over 100 million implantations worldwide) are known to develop significant glistenings (microvacuoles) and whitenings (sub-surface nanoglistenings) over time (24)(25). At three years, around 90 percent of patients will have glistenings on their implanted IOL, with moderate to severe grade glistenings occurring in over 60 percent of patients (26)(27)(28)(29)(30)(31). One study has suggested that glistening-induced light scattering can continue to increase in AcrySof lenses for 15 years post-operatively (32). One study has even shown that five years after AcrySof implantation, 43 percent of patients reported difficulties with night driving and modified their driving habits accordingly (33). Our own research has indicated that AcrySof-implanted subjects are more likely to have self-reported road traffic accidents at five years, and worse visual acuity, than patients implanted with Tecnis IOLs (Table 3). What’s more, evidence suggests that removing the AcrySof lens ameliorates the increased risk of RD, disability glare and IOL dislocation; our data from five patients has shown that visual acuity is improved when IOLs with severe glistening and whitening are exchanged for clear IOLs (24). Similarly, explantation of multifocal acrylic IOLs with glistenings in the optic, and replacement with monofocal IOLs, resolved symptoms and improved visual acuity by up to 4 lines (36).

Conclusions

My strong feeling is that it is time to rethink the RLE procedure. I recognize that I may be ‘swimming against the tide’ – but I am not swimming alone. The problems with RLE have started to become apparent to those who think about these matters. For example, this very issue was raised last year by a colleague in Mexico who frequently participates in clinical trials involving innovative lenses, and therefore sees results that may not be well-publicized. It is interesting that Alcon has very recently introduced a glistening-free lens; perhaps this could be read as an acknowledgement of the problems associated with the current generation of IOLs.  

In summary, my basic premise is that RLE is not as safe or effective as we believe, and we need to collectively sit back, look at all the data and think again about how it is used. I am not against IOL procedures altogether – but I certainly try to dissuade young patients with clear lenses from electing for RLE. I show them all the data and tell them about the potential problems, and try to persuade them to try contact lenses or even corneal surgery instead. My experience has been that most patients, when presented with the data, decide against the RLE procedure – and I think that’s wise.