Official ESCRS | European Society of Cataract & Refractive Surgeons


Implantable collamer lens (ICL) vault and lens separation prediction using a multivariate regression model based on parameters acquired by very high-frequency digital ultrasound

Search Title by author or title

Session Details

Session Title: Posterior Chamber Phakic IOLs

Session Date/Time: Sunday 15/09/2019 | 14:00-16:00

Paper Time: 14:56

Venue: Free Paper Forum: Podium 3

First Author: : D.Reinstein UK

Co Author(s): :    R. Vida   T. Archer                          

Abstract Details


To derive a model to predict vault after Implantable Collamer Lens (ICL) surgery using very high-frequency (VHF) digital ultrasound biometry


London Vision Clinic, London, UK


Retrospective analysis of 42 consecutive V4c ICL procedures (STAAR Surgical). Initial ICL size was chosen based on VHF digital ultrasound imaging, with reference to the Dougherty and Kojima published nomograms. Stepwise multivariate regression analysis was performed to predict central vault using the following variables: ICL size, ICL power, sulcus-to-sulcus (STS), ciliary body diameter (CBD), zonule-to-zonule, crystalline lens rise from the sulcus plane (STSL) and zonular plane, anterior chamber depth, anterior chamber angle, scotopic pupil diameter (SPD), angle-to-angle, and white-to-white. The regression model was applied to compare vault predictability of current ICL sizing formulae and a theoretical 12.9mm size.


Significant variables were ICL size, ICL power, CBD, STSL, and SPD. In comparison to the model our current population had the same lens selected in 43%, one size different in 53% and 2 sizes different in 6%. STAAR calculator: 35%, 41%, and 23%, respectively. Dougherty formula: 43%, 56%, and 3%, respectively. Using an ideal lens size (closest to achieving a 500µm vault), a vault of 496±71µm (384 to 611µm) would be predicted. Had a 12.9mm lens been available, this would have been selected in 35%, predicting a vault of 490±51µm (404 to 589µm).


The ciliary body diameter proved to be the most significant lateral measurement, even over STS. The regression model also found pupil diameter to be significant which has not been a part of any previously published model. The model showed that using the STAAR calculator and white-to-white would greatly increase the chance of choosing a lens that is too large. Finally, the model showed that the vault and standard deviation could be greatly improved if a 12.9 lens size was available.

Financial Disclosure:

has significant investment interest in a company producing, developing or supplying product or procedure presented

Back to previous