The Role of Corneal Biomechanics for the Evaluation of Ectasia Patients

Abstract

Purpose: To review the role of corneal biomechanics for the clinical evaluation of patients with ectatic corneal diseases.

Methods: A total of 1295 eyes were included for analysis in this study. The normal healthy group (group N) included one eye randomly selected from 736 patients with healthy corneas, the keratoconus group (group KC) included one eye randomly selected from 321 patients with keratoconus. The 113 nonoperated ectatic eyes from 125 patients with very asymmetric ectasia (group VAE-E), whose fellow eyes presented relatively normal topography (group VAE-NT), were also included. The parameters from corneal tomography and biomechanics were obtained using the Pentacam HR and Corvis ST (Oculus Optikgeräte GmbH, Wetzlar, Germany). The accuracies of the tested variables for distinguishing all cases (KC, VAE-E, and VAE-NT), for detecting clinical ectasia (KC + VAE-E) and for identifying abnormalities among the VAE-NT, were investigated. A comparison was performed considering the areas under the receiver operating characteristic curve (AUC; DeLong's method).

Results: Considering all cases (KC, VAE-E, and VAE-NT), the AUC of the tomographic-biomechanical parameter (TBI) was 0.992, which was statistically higher than all individual parameters (DeLong's; p < 0.05): PRFI- Pentacam Random Forest Index (0.982), BAD-D- Belin -Ambrosio D value (0.959), CBI -corneal biomechanical index (0.91), and IS Abs- Inferior-superior value (0.91). The AUC of the TBI for detecting clinical ectasia (KC + VAE-E) was 0.999, and this was again statistically higher than all parameters (DeLong's; p < 0.05): PRFI (0.996), BAD-D (0.995), CBI (0.949), and IS Abs (0.977). Considering the VAE-NT group, the AUC of the TBI was 0.966, which was also statistically higher than all parameters (DeLong's; p < 0.05): PRFI (0.934), BAD- D (0.834), CBI (0.774), and IS Abs (0.677).

Conclusions: Corneal biomechanical data enhances the evaluation of patients with corneal ectasia and meaningfully adds to the multimodal diagnostic armamentarium. The integration of biomechanical data and corneal tomography with artificial intelligence data augments the sensitivity and specificity for screening and enhancing early diagnosis. Besides, corneal biomechanics may be relevant for determining the prognosis and staging the disease.

Keywords: corneal biomechanics; corneal ectasia; corneal tomography; keratoconus.

Figure 1

Receiver operating characteristic (ROC) curves from tomographic-biomechanical parameter (TBI), pentacam random forest index (PRFI), Belin–Ambrosio Display D value (BAD-D), corneal biomechanical index (CBI), Pentacam Topographic Keratoconus classification index (TKC) and IS Abs considering groups 2, 3, and 4. Note that the area under the curve AUC of the TBI was statistically higher than all other parameters.

Figure 2

AUSEP- area under the superation curves from TBI, PRFI, BAD-D, CBI, and IS-Abs considering groups 2, 3, and 4. Note the higher spread among the metrics with the TBI.

Figure 3

ROC curves from TBI, PRFI, BAD-D, CBI, TKC, and IS Abs considering only groups 2 and 3. Note that TBI had almost 100% sensitivity to detect frank ectasia (AUC 0.999), being once more statistically higher than all other parameters described.

Figure 4

AUSEP curves from TBI, PRFI, BAD-D, CBI, and IS-Abs considering only groups 2 and 3. Note, once more, the higher spread among the metrics with the TBI.

Figure 5

ROC curves from TBI, PRFI, BAD-D, CBI, TKC, and IS Abs considering the with very asymmetric ectasia with relatively normal topography (VAE-NT) group. Note that the AUC of the TBI was 0.966, which was once more statistically higher than all other parameters tested.

Figure 6

AUSEP curves from TBI, PRFI, BAD-D, CBI, and IS-Abs considering only groups VAE-NT group. Note, once more, the higher spread among the metrics with the TBI.

Figure 7

Graphical representation of the data distribution-Dot plots. Note one more time the better performance of TBI compared to PRFI, BAD-D, and CBI.

Figure 8

(A) Tomographical biomechanical display (ARV) from the right eye. Note, on the Pentacam Tomographic Assessment (top right), that the front surface curvature demonstrates an advanced KC condition on this eye. (B) Tomographical biomechanical display (ARV) from the left eye. Note a relatively normal anterior curvature map on the Pentacam Tomographic Assessment (top right). Interestingly, deformation corneal response revealed an abnormal TBI value of 0.70 in the left eye, even though CBI (0.00) and BAD-D presented normal values.

Figure 9

(A) and (C): Anterior curvature maps from OD in 2019 and 2017, respectively. (B) and (D): Anterior curvature maps from OS in 2019 and 2017, respectively. The right-hand panel demonstrates the differential maps from OD, which was the eye that presented the highest progression.

Figure 10

Biomechanical comparison displayed from the right eye in 2017 (Exam A) and 2019 (Exam B). Note that the cornea becomes softer and thinner, considering the first and second consultations, as noted comparing the variables DA ratio, integrated radius, SP A1, and ART h.

Figure 11

Slit-lamp biomicroscopy photo from the right eye. Note the hexagonal corneal incisions.

Figure 12

Placido disk-based corneal topography from OD and OS demonstrating the ectatic condition in OU.

Figure 13

Vinciguerra Screening Report from the right eye evidencing abnormal biomechanical parameters, even after adjusting the software for post-laser-vision-correction (LVC) state. Note the abonormal CBI value.