Why single-instrument screening for keratoconus remains a dream

The promise of simple, objective, single instrument-based screening for highly asymmetric keratoconus in patients considering LASIK is still more of an idea than reality. Multiple metrics can identify early keratoconus in a minority of affected eyes, but none of the metrics consistently identify early corneal abnormalities and none of the metrics identify the same eyes as being at elevated risk for ectatic disease.

“We know that the risk of ectasia is highest in patients with some predisposition to keratoconus,” said J. Bradley Randleman, MD, professor of ophthalmology, University of Southern California Keck School of Medicine and director of cornea and refractive surgery at the USC Roski Eye Institute, Los Angeles, CA. “In patients with highly asymmetric keratoconus, the less affected eye should, in theory, be the one to be able to test imaging modalities and develop simple screening criteria. The problem is that none of the metrics we have and none of the instruments we have, do a particularly good job in isolation of distinguishing between normal eyes and the less affected eye with great precision.”

Dr. Randleman, who is also Editor-in-Chief of The Journal of Refractive Surgery, presented an analysis of metrics derived solely from Scheimpflug imaging (Pentacam HR, Oculus) for the detection in highly asymmetric eyes with keratoconus.

LASIK is one of the most successful optical surgeries, but success is dependent on good patient selection. Keratoconus can dramatically increase the risk of ectatic disease following refractive surgery. A minority of patients with keratoconus present with what is traditionally called unilateral keratoconus, or pronounced disease in one eye only.

But “unilateral” is misleading, Dr. Randleman explained, as patients rarely have keratoconus in only one eye. Like many ophthalmologists, he prefers the term highly asymmetric keratoconus as a more accurate description. Keratoconus is usually easy enough to determine when patients have obvious signs of keratoconus and avoid LASIK, but the subtler, initial findings are the ones still in discussion and the ones that make screening challenging. 

Clinicians traditionally assess these questionable eyes using a subjective combination of corneal measurements and clinical experience. A global consensus published in 2015 (Gomes et al. Cornea. 2015;34(4):359-369) reported that tomography was the best way to screen for early or subclinical keratoconus and that posterior corneal elevation abnormalities must be present. Many authorities take exception to these conclusions, citing a lack of evidence. 

“The results of our study specifically argue against the opinions presented in the Global Consensus paper, as the metrics they identified as being required were in fact not found in the majority of our patient population,” Dr. Randleman said.

The problem is that each instrument currently on the market evaluates different metrics. Ophthalmologists might prefer one instrument or another, or use multiple instruments to assess questionable eyes, but the decision to perform refractive surgery remains a highly subjective process.

“Most eyes are appropriate for LASIK, but you frequently want more information on up to 10 % of eyes you see on initial screening,” Dr. Randleman said. “At this point, subjective analysis remains absolutely critical for good results. That is why it is imperative the surgeons continue learning what is normal and what is suspicious, and keep up to date with the current literature. And for now, you have to rely on different metrics from multiple instruments.”           

The Pentacam HR has been considered by some to be the most useful instrument to diagnose early keratoconus using Scheimpflug imaging, Dr. Randleman said, but there has been little evidence to support that assumption.

He presented the results of a study comparing 25 clinically normal eyes of patients with definitive keratoconus in the contralateral eye with 50 eyes from 50 patients that had normal evaluations by multiple imaging devices with uneventful LASIK and at least one year of follow-up.

The study evaluated anterior curvature, thickness, topometric maps, Zernike maps, and the enhanced ectasia map. The goal was to evaluate objective metrics derived from Pentacam Scheimpflug imaging and determine their relative value in distinguishing between normal control eyes and the less affected eye of patients with high asymmetric keratoconus.

No metrics achieved excellent or good scores. Three metrics were scored fair in distinguishing suspect eyes from normal eyes, ISV keratoconic index from the topometric map, and the D score and ARTmax from the enhanced ectasia map. None of the three metrics were particularly useful individually or in combination, Dr. Randleman said. ISV identified 20 % of suspect eyes, D score 12 % and ARTmax 24 %, but the three metrics all identified different eyes.

“The final message is that you cannot rely solely on machine-based metrics at this time,” Dr. Randleman said. “The goal is to transform the subjective nature of screening into an objective evaluation, but we’re not there yet. When you are screening for refractive surgery, until we have one metric that is fully reliable, you still have to err on the side of caution. When we see patients who are not good candidates for LASIK, we need to be extremely cautious.”

Patients with any identifiable predisposition for keratoconus should not have surgery in either eye, Dr. Randleman said.

“We know what full blown keratoconus looks like, but the subtler forms are the ones that make screening challenging,” he said. “That is why we are looking at these highly asymmetric patients-we want to see what metrics we can apply for screening in those eyes to other patients with no obvious findings in either eye.” 

This article was adapted from Dr. Randleman’s presentation at the 2016 American Society of Cataract and Refractive Surgery Symposium & Congress.