La ligne de démarcation stromale est-elle prédictive de l'efficacité du Cross-Linking.
LHUILLIER L*, HOUMAD N (Metz), ZAIDI M (Nancy), SOT M, EHRHARDT A, KRAWCZYK P, GUECHI O, AGAPIE A (Metz), DUONG
NGUYEN VIET H (Hô Chi Minh-Ville, Viêt Nam), PERONE JM (Metz)
Introduction : To assess the association between stromal demarcation line after collagen crosslinking (CXL) for
progressive keratoconus and mid-term (6 month) refractive, topographic and clinical outcomes of CXL.
Matériels et Méthodes : All consecutive patients treated with CXL between March 2014 and March 2015 were
included. They all underwent the same procedure: accelerated-CXL protocol (10 minutes UVA irradiation with 9
mW/cm² intensity), irradiation device (IROC UV-X™ 2000, Medeuronet). Iso-osmolar riboflavin solution was used for
patientswith corneal thickness above 450 µm (thinnest point before de-epithelization),hypo-osmolar solution was used
for thinner corneas (400-450 µm before de-epithelization). One month after the procedure, AS-OCT examination was
performed in all patients (RS-3000, NIDEK). The visibility of the stromal demarcation line was scored (0: not visible
line; 1: visible line, but measurement not clearly defined; 2: clearly visible line) and its depth was measured centrally at
anytime it was possible. Changes in best corrected visual acuity (BCVA), spherical equivalent (SE), Kmax and thinnest
pachymetry, were studied preoperatively and at 6 months.
Résultats : 36 eyes (29 patients) with progressive keratoconus were included mean age was 26.1 ± 10.4 years. One
month after CXL, a stromal demarcation line was visible for 28 patients (77.8%) (grade 2 : 15, 41.7% and grade 1 : 13,
36.1%) and not visible ( grade 0) for 8 (22.2%) patients. Mean depth of the corneal stromal demarcation line was 342.3
± 58.5 μm (range 226-440μm). Six month after CXL, the study showed : a significant improvement in BCVA (
preoperative : 0.5LogMar ±0.2 vs postoperative 0.39LogMar ±0.2 , p = 0.031) and SE ( preoperative -6.46 D ±2.1 vs
postoperative 5.42D± 2.1 , p = 0.027) ; a significant decrease in thinnest pachymetry ( preoperative : 451.6 µm ± 30.8
vs postoperative 421µm ± 46.9, p = 0.043). No significant change for Kmax (preoperative: 58.1 D ± 5.4 vs postoperative
57.8 ± 4.5, p=0.35). Neither the visibility of the stromal line nor its depth was significantly associated with post-
operative changes in Kmax, SE, BVCA or pachymetry (p>0.05). This study tends to show that preoperatively stepper
and thinner corneas were associated with a more visible line: Kmax = 58.3 ± 5.5D (visible) vs 54.5 ± 5.8D (not visible
line) p=0.051 ; thinnest pachymetry = 451.2 ±34.9µm (visible) vs 477.1 ± 26.2 (no visible line) p= 0.059.
Discussion : Preoperative corneal characteristics might be associated with the demarcation line since stepper and
thinner corneas trend to have more visible line.
Conclusion : In our study neither the visibility of the stromal line nor its depth was significantly associated with post-
operative changes in Kmax, SE, BVCA or pachymetry (p>0.05). The stromal demarcation line was not predictive for CXL
efficiency 6 month after the procedure.
La ligne de démarcation stromale est-elle prédictive de l'efficacité du Cross-Linking.