Playing Our Cards Right

If percentage tissue altered (PTA – the combination of the flap thickness plus the ablation depth divided by the pre-operative central corneal thickness) is high, it represents a risk factor for post-LASIK ectasia (

)()()()()(). And the higher the PTA, the higher the risk. I’d like to highlight three important points to go along with that statement: firstly, the concept of PTA comes from a solid theoretical foundation. Secondly, it is a risk factor and not a screening method. And thirdly, how risk factors are investigated is of utmost importance.

The creation of a LASIK anterior lamellar flap should not normally be associated with a significant loss in corneal biomechanical strength. However, corneal tensile strength is not uniform throughout the central cornea (posterior corneal stromal tissue is weaker than anterior stromal tissue – especially the posterior two-thirds of the cornea), meaning that the deeper the LASIK flap cut, and the greater the amount of tissue ablated, the weaker the remaining cornea becomes (

). Based on these structural differences, it is reasonable that a ratio or equation would be representative of post-LASIK changes, specifically of values of residual stromal bed or corneal thickness. And that’s why we proposed measuring PTA as a risk factor for post-LASIK corneal ectasia.

When we first started trying to determine new methods of assessing corneal biomechanics with BJ Dupps, we also looked at which intraoperative variables induced changes in these parameters. PTA was one of those investigated and, to our surprise, it had the highest number of significant correlations with changes in the biomechanical variables under assessment. As ectasia likely represents a reduction in biomechanical integrity below the threshold required to maintain corneal shape and curvature, we wanted to understand if there was an association between PTA and ectasia. We identified that PTA was significantly higher in a group of patients who developed ectasia after LASIK compared with a group of patients who hadn’t developed any complications three years after surgery (

). We then proceeded to investigate ectatic patients who had normal pre-operative topography in a case-control study, and identified that PTA equal to or higher than 40 percent was by far the most prevalent risk factor, and had an odds ratio of 223 ()!

Our case-control study was appropriately designed to investigate PTA as a risk factor – and this matters. Saad et al. (

) also investigated PTA in a retrospective cohort study of 126 eyes with PTA ≥40 percent and an average of two years follow-up after LASIK, but did not identify it as a risk factor for post-LASIK ectasia. Though a good paper in some regards, some serious issues should be highlighted. One huge flaw is that the outcome of ectasia was not present in any of the participants. The outcome is needed to investigate any risk factor – it’s like I am investigating mortality and nobody died! It is methodologically wrong to draw a conclusion about a risk factor from a population that didn’t develop the adverse event. Remember, people who have high a PTA might not develop ectasia – it is a risk factor not a predictor. The most important risk factor, for death is high blood pressure (). If you’ve got high blood pressure, don’t worry, it doesn’t mean you will die; it just means that your chance of dying is a little bit higher compared with someone in the normal blood pressure range.

Another flaw was the actual design of the study; cohort studies are for common outcomes with a rare risk factor. Ectasia is not a common outcome – the authors themselves cited it as having an incidence of 0.04–0.6 percent (

). With such a low incidence, a huge number of participants would be needed to draw any conclusions – yet they only had 126. Calculating the sample size that would be needed for this kind of study, based on an incidence of 0.6, you’d need around 3,500 participants with the risk factor. Based on an incidence of 0.04, approximately 53,000 participants would be needed. Using the incidence of 1/2,500 (0.0004 percent) as cited by Randleman et al. (), over 5 million participants would be needed – that’s why cohort studies aren’t used to investigate risk factors. Case control studies, like ours (), should be used for rare outcomes with common risk factors. Furthermore, the patients in the Saad et al. study were only followed up for an average of two years, but it has been shown that only 20 percent of ectasia cases are detected two years after surgery; this jumps to 50 percent at four years and 75 percent at seven years ().

We’ve not finished our work on PTA as a risk factor for ectasia – we’ll shortly be publishing our tenth paper. We understand that flap thickness is different to ablation depth so we are working on an equation that gives constants for those variables. There will be people with a PTA higher than 40 percent who will never develop ectasia, because ectasia is a rare adverse event. But our take home message is very simple: high PTA is a risk factor for ectasia. And we’ve proved it through a correctly-designed case-control study that demonstrated a high odds ratio (

).