The dangers of Lasik

It is surely time to warn both eye care professionals and the public of the dangers of laser in situ keratomileusis (Lasik). The corneal flap of approximately 16µm, of one-third thickness of the average cornea, has been shown to never heal properly by Seiler (Dresden) and Marshall (London). Approximately 22 million corneal fibres are intersected, their severed ends never rejoining, meaning that the flap is only held in place by glycosaminoglyans. To put it more simply, the corneal flap after Lasik provides no more corneal strength than the wearing of a contact lens. Marshall has also reported that studies using high frequency imaging have shown that the Lasik flap interface moves during accommodation.
Vapourisation of corneal tissue (ablation) by the laser removes a precise depth of stroma, 50µm for a -4.00D correction, 120µm for a -7.00D correction. Simple mathematics tell us that with a flap of 160 microns and an average corneal thickness of 550µm, the cornea is weakened by 38 per cent in a -4.00D correction and 51 per cent in a -7.00D correction.
Furthermore, the corneal weakening which occurs in every case of Lasik is probably even more severe. Bron (Oxford) has demonstrated that the anterior 100-120µm of the cornea (which are included in the Lasik flap) have the greatest strength and Muller (Holland) has shown that the anterior cornea has the greatest stability.
Aware of these facts, Lasik surgeons are measuring corneal thickness pre-operatively and sometimes inter-operatively, estimating that a safe residual corneal thickness under the ablation bed of at least 250µm should remain. Lasik manufacturers have warned against high corrections of more than -7.00D and the exclusion of thin corneas. Studies by Amoils (Johannesburg) demonstrate that this does not solve the problem.
In the face of great popularity among patients rejoicing in almost painless, instant restoration of vision, massive advertising and large financial returns, the concerns of a few refractive surgeons are being ignored.
In 1998 Amoils reported a series of cases of keratoconus (KC) occurring in post-Lasik patients in previous normal corneas with no incipient KC or 'form fruste' corneas.
It is noteworthy that Pallikaris (Crete), the pioneer of Lasik in 1990, is now performing epi-Lasik, citing biomechanical problems with Lasik. In epi-Lasik, only a very thin corneal epithelial flap is raised. O'Brart (London) also aware of these problems, performs Lasek in which a much thinner corneal flap is raised.
The American Society of Cataract and Refractive Surgery (ASCRS), the largest of its kind in the world, is sufficiently concerned that it is polling its entire membership to report cases of corneal ectasia, or forward bowing, of the posterior corneal surface, that can be an early sign of KC.
Marshall has observed eyes that have needed corneal grafts due to late atrophic changes in the corneal flap. He expects that in the long term Lasik patients will experience a shift out of focus again due to the biomechanical effect of the flap.
As long ago as December 1994, the editor of the European Journal of Implant and Refractive Surgery (Emanuel Rosen) in his editorial stated 'the concept of treating myopia by significant corneal thinning flies in the face of clinical wisdom'.
Excimer laser corneal ablation can also cause loss of contrast sensitivity in a high proportion of cases. This significantly reduces night vision and can impair safe night-time driving. This has been reported in several university studies. In 1996 TŸbingen University (Germany) at the ASCRS meeting in Seattle announced that, over a 10-year period, 75 per cent of their patients having excimer laser treatment for myopia were no longer legally able to drive at night in Germany.
We at the London Centre for Refractive Surgery, therefore, recalled our patients and at the Boston ASCRS meeting in 1998 reported that 56 per cent of them after PRK had reduced contrast sensitivity. We therefore abandoned the use of the excimer laser. The following year, Ragensburg University (Germany) reported a 53 per cent incidence of contrast sensitivity loss. In response to my paper, Ottawa University reported to me that they had similar figures and that their findings were not welcomed by the laser industry. However, the federal government in Ottawa advised all provincial governments to append a night vision test as part of the driving licence criteria.
In March 2001 at an international refractive surgery symposium in London, chaired by George Waring III, Moorfields Eye Hospital in London, reported that reduced contrast sensitivity was found in 30 per cent of their patients two years post-PRK and in 50 per cent of their patients one year post-Lasik. Pupil size had not been recorded, but there is conflicting evidence of whether this makes more than a small difference. However, Rosen (Manchester) disagrees and has shown studies that preoperative pupil measurement is important to exclude night vision problems. By comparison, our results after incisional keratotomy, ignoring pupil size, showed a 3 per cent incidence of contrast sensitivity loss. Despite attempts to have these findings published, none were successful in seeing their figures reported in peer-reviewed journals. The chief reason given was that there were no peer-reviewed articles to which they could refer on the same subject, so that we were all in a Catch-22 situation.
However, in June 2002, Semmelweis University (Budapest) and the Cleveland Clinic in Ohio managed to report in a peer-reviewed journal that 18.5 per cent of their patients with a flying spot laser PRK and 31.6 per cent with scanning beam laser PRK had significantly reduced contrast sensitivity, despite minimal corneal haze being present.
Although I agree with Rosen that pupil size is important, the mechanism of this loss of contrast sensitivity is almost certainly the penetration of some of the laser's energy beneath the bed of ablation, disturbing the perfect symmetry of the corneal lamellae and their interspaces. Apart from a corneal transplant, there is no cure for this.
Wavefront technology has been widely advertised and promoted as giving greatly improved results with the excimer laser. It is certainly a major advance in the separation and analysis of refractive errors within the eye. It definitely reduces corneal aberrations induced by the laser. But, however precise its measurements are pre-operatively, they are to a great extent negated by the corneal cut in Lasik and the disturbance of the perfect corneal symmetry caused by ablation.
Several very sophisticated instruments can be used to measure contrast sensitivity, but the Pelli Robson chart gives reliably reproducible results.
What else is available to the refractive surgeon? A few centres still perform radial keratotomy (RK)or incisional microsurgery. It has its disadvantages, notably a lack of 100 per cent predictability (a fault shared with the laser) due to differing healing characteristics, a 20 per cent incidence of longer-term influence causing hyperopia (although by conservative surgery we have reduced this to 0.4 per cent incidence of secondary hyperopia in a series of 10,000 consecutive RK microsurgeries for myopia). Olsen (US) has measured reduction of corneal tensile strength by up to 10 per cent, but none of our cases over 17 years has developed keratoconus. Robin (US) showed that the incidence of traumatic rupture after RK is no greater than in the normal population.
Critics often deride RK microsurgery as out of date, but very few of them have personal experience of doing this operation. It is ironic that arcuate keratotomy is widely performed to correct post cataract and post corneal graft astigmatism, but the biomechanics of curved corneal incisions are no different from straight radial ones.
With its much longer safety record we recommend RK microsurgery for lower myopic corrects, being the instructors for the ESCRS surgical skills courses.
Nevertheless, changing the shape of the cornea, especially in myopes, by flattening the centre and increasing the peripheral curvature or changing corneal shape from prolate to oblate, impairs natural vision even if acuity is not lost. Holladay (Texas) has memorably stated, 'the eye of the hunter becomes that of the hunted'. There is a general recognition emerging that phakic lens implants are the most natural way of correcting refractive errors. Placing these posterior to the pupil, just clearing the anterior lenticular surface, is theoretically ideal, but can interfere with aqueous nutrition of the crystalline lens, leading to its opacification. For this reason, most surgeons prefer to place the implants in the anterior chamber.
Phakic implants are generally made of PMMA (Perspex) rather than acrylic or silicone, which can opacify. These lenses are available heparin coated to reduce the risk of anterior uveitis.
Currently, phakic lenses are implanted through a 5.0mm clear corneal incision, necessitating corneal suturing. However, we now have foldable lenses, which are introduced through a 4mm incision and in the near future expect to inject such lenses through a 1.5mm incision, avoiding the requirement for corneal suturing.
Since they lie in front of the pupillary aperture, they can interfere with aqueous flow, therefore a peripheral iridectomy is mandatory to avoid the risk of acute glaucoma.

progress to date
To date, approximately 250,000 phakic lenses have been implanted worldwide.
Provided careful assessment of preoperative anterior depth is made, posterior corneal touch, which can lead to corneal decompensation and corneal clouding, is avoided, even in 75 per cent of hyperopic patients.
Nevertheless, it is important to measure corneal endothelial cell count with a specular microscope. Precise sizing of the implant, accurate localisation of its optic centrally and correct placement of its haptics, or supporting limbs, in the angle of the anterior chamber are essential to avoid pupil ovalisation and achieve optimal visual acuity.
Since the globe has to be opened, the consequences of wound infection are more serious, the current risk being reported at between 1 in 1,000 and 1 in 5,000 cases. Likewise, the possibility of subsequent retinal detachment, especially in highly myopic eyes, although small, is increased.
Co-management of these cases by local eye care professionals is important, particularly in the early postoperative period.
Currently, we use phakic 6 lenses which have a corrective range of -25.00 to +8.00D, made of PMMA and are the only ones heparin coated. Also available are Nuvita to -23.00D, Artisan which is clipped to the iris with claws and Staar lenses which are placed in the posterior chamber.
With these lenses, a very high range of refractive errors of myopia, hyperopia and astigmatism can be permanently corrected. Despite their inherent accuracy, postoperative astigmatism can still occur, which is most safely corrected by astigmatic keratotomy.
Lens implants can be made bifocal for presbyopia, but this is not true presbyopic correction. The use of small PMMA implants inserted into the four quadrants of the anterior scleral coat slightly contract the ciliary body by compression, allowing it to act more efficiently and this appears to be a promising surgical procedure for the correction of presbyopia.
The concept of a 'SMART lens' was discussed at the ASCRS meeting in San Francisco in April 2003. In these cases, a normal lens extraction would be done by zero energy aspiration through a 1.5mm clear corneal incision and a cylinder of an advanced polymer containing memory of the patients' distance correction would be placed in the capsule. Body temperature and hydration would convert this into a normal lens shape. The lens would have memory, so distance glasses would not be required, and, since the lens material would remain flexible throughout life, presbyopia would be 'avoided'. Since the normal crystalline lens would have been removed, cataract formation would never occur. Perhaps this would be the answer to cataract waiting lists, but this tantalising prospect lies in the future.

William Jory is a consultant ophthalmologist based in Harley Street. optician will publish a counter-argument by Professor Emanuel Rosen in the coming weeks