Discerning between high astigmatism, subclinical keratoconus is problematic

Reviewed by Maria A. Henriquez, MD, MSc, PhD

Currently available keratoconus screening indices found on two commercially available Scheimpflug devices show a lack of predictability for accurately discriminating between subclinical keratoconus and normal eyes with high astigmatism, according to findings of a retrospective study conducted by Maria A. Henriquez, MD, MSc, PhD, and colleagues.

“The screening indices found on existing topography or tomography systems perform well for discriminating between normal eyes and eyes with keratoconus. However, discriminating between low ametropia and clear keratoconus is a different situation than discriminating between eyes with high ametropia versus subclinical keratoconus,” said Dr. Henriquez, director, Research Department, Oftalmosalud Instituto de Ojos, Lima, Peru, and professor of ophthalmology, Post GCRST Harvard Medical School, Boston.

“Our results show that eyes with high astigmatism are very often falsely diagnosed as having subclinical keratoconus using the keratoconus screening indices,” Dr. Henriquez said. “This creates a problem for surgeons because they are faced with trying to decide whether to operate on a patient whose parameters fit within the abnormal values or, because of the doubt raised by the machine’s results, to not operate on a patient whose eye may be normal.”
The study included 67 eyes of 67 patients who were evaluated with two Scheimpflug imaging devices (Pentacam, Oculus; Galilei, Ziemer) prior to undergoing LASIK. All eyes had >1.5 D of astigmatism in the subjective refraction and anterior cornea astigmatism and had been followed for at least 12 months after LASIK without evidence of developing ectasia.

The accuracy of 10 preoperative keratoconus screening indices found on each device and/or suggested by previous studies for diagnosing both keratoconus and subclinical keratoconus was evaluated by analyzing its rate of false positives in the sample size (i.e., eyes that were wrongly identified as having keratoconus or subclinical keratoconus).

“The indices and cutoff point we used represented those that had been suggested by previous studies as having the highest predictive value for identifying keratoconus and subclinical keratoconus from normal eyes,” Dr. Henriquez explained.

In the analysis of diagnostic performance for correctly identifying keratoconus, the false positive rates for all of the Pentacam screening indices except for one index fell within the previously suggested range. However, with the Galilei, only 2 of the 10 indices, had a false positive rate that did not exceed the previously reported maximum.

For diagnosing subclinical keratoconus, 6 of the 10 Pentacam screening indices and 7 of the 10 Galilei indices were associated with false positive rates that exceeded the values suggested as acceptable in previous studies or by the manufacturers.

Subgroup analyses were performed to determine the false positive rates for diagnosing keratoconus and subclinical keratoconus with eyes categorized by type of astigmatism (myopic, hyperopic, mixed, or simple). Overall the results showed that false positive rates were highest in the hyperopic astigmatism and mixed astigmatism groups, which leads us to think that there is still a need for better diagnostic tools for this subgroup of patients.

Disclosures:

Maria A. Henriquez, MD, MSc, PhDE: mariale_1610@hotmail.com
This article is based on a presentation given by Dr. Henriquez at the 2018 meeting of the American Society of Cataract and Refractive Surgery. She has no relevant financial interest to disclose.