Early influence of endotamponade on corneal biomechanical parameters, central corneal thickness and accuracy of intraocular pressure measurement

To define the influence of air, SF6 gas and silicon oil 1000cs tamponade injection and oil tamponade removal on corneal biomechanics, central corneal thickness and intraocular pressure. 77 eyes referred to vitrectomy were divided into 4 groups: 19 to air tamponade, 21 to SF6 tamponade, 19 to oil tamponade, 18 to oil tamponade removal. Pre- and postoperative corneal hysteresis, corneal resistance factor, corneal thickness, Goldman intraocular pressure (GAT) and corneal compensated intraocular pressure (IOPcc) were analysed. GAT and IOPcc did not change after the air or SF6 tamponade. The oil tamponade injection caused increase in GAT and IOPcc, while the oil removal caused reduction in those parameters. In all groups, preoperative and postoperative values of GAT and IOPcc did not differ. There was no change in corneal thickness and biomechanics after air, SF6 or oil tamponade while after removal of oil those parameters are decreased. The air, SF6 and oil tamponade does not change the corneal thickness and corneal biomechanics. The removal of oil causes decrease in corneal thickness and biomechanics which can lead to bias in intraocular pressure measurement. GAT and IOPcc did not differ in eyes pre- and post-vitrectomy, being similarly reliable measure.


Results
In the first study group (patients referred to PPV with intravitreal air application) there were statistically significant differences in mean values of: CCT (t = − 3.11, p < 0.05) and WS (t = 5.63, p < 0.05) when compared the pre and postoperative measurement taken on 1st postoperative day (Fig. 1).

Discussion
The common use of vitreous substitutes such as: air, gas (SF6, C3F8), silicon oil (1000cs, 2000cs, 5000cs) tamponade improve the PPV outcomes, however it is also related to some postoperative complications. One of the most severe complications following the PPV surgery with tamponade injection is increase in IOP, which can lead to mechanical compression and blood flow disturbances within the optic nerve head. The damage to retinal  www.nature.com/scientificreports/ nerve fiber layer (RNFL) and ganglion cell complex (GCC) results in irreversible glaucomatous changes followed by visual field defects [2][3][4][5] . In addition, the use of intravitreal substitutes can cause corneal changes such as: loss of corneal endothelial cells, corneal decompensation, band keratopathy and can induce postoperative changes in CCT and corneal biomechanical parameters as CH and CRF [6][7][8][9][10][11] . This changes, in turn, influence the accurate IOP measurement even with the leading tonometry methods.
In the study of Muether et al., in eyes after PPV with SF6 tamponade, the early significant rise in GAT (measured on the first postoperative morning) was noted in 37.9% of cases 12 . Our results strongly confirmed the Muether's reports, as there was significant rise in GAT and IOPcc on 1st postoperative day in patients with SF6 tamponade when compared to preoperative measurements. In contrast, in the group where the air tamponade was applied, there was no significant difference in GAT or IOPcc when measured on the 1st day after the surgery. In the study of Teke M, who compared IOP values measured with ORA (IOPg and IOPcc) before and 1 week after the PPV with SF6 or C3F8 injection, there was significant rise of IOPcc and IOPg in group with long acting C3F8 gas, with no significant change of IOPcc and IOPg values in SF6 group 7 . This is consistent with our results, as there was no significant rise in GAT and IOPcc in both groups of patients, who underwent PPV with intravitreal air or SF6 injection when measured on 7th postoperative day. It is thought that the reason for lack of change in IOP in patients with air or SF6 injection 7 days after the surgery is lack of air expansile properties and slightly expansile concentration of SF6 (25%) which does not elevate IOP significantly. Table 3. Correlations of selected parameters assessed pre and postoperatively in the third study grouppatients with RRD qualified for PPV with 1000cs silicone oil injection (GAT Goldman intraocular pressure, IOPcc-corneal compensated intraocular pressure, CH corneal hysteresis, CRF corneal resistance factor, CCT central corneal thickness, pre measured on the 1st preoperative day, post 1 measured on the 1st postoperative day, post 7 measured on the 7th postoperative day).    www.nature.com/scientificreports/ after silicon oil tamponade removal 16 . Moreover, lots of authors described significant intraocular hypotony, defined as IOP values of 5 mmHg or less after the silicon oil removal [17][18][19][20] . However, early intraocular hypotony was not observed in our study as the mean values of GAT and IOPcc were both above 9.0 mm Hg (respectively: 9.01 ± 3.6 SD, 9.33 ± 3.08 SD) when measured on 1st day postoperatively. In all groups enrolled in our study, the preoperative and postoperative values of GAT and IOPcc did not differ significantly. Moreover, our study showed strong positive correlation between GAT and IOPcc what can suggest that both measurements are similarly reliable in patients who have undergone PPV with tamponade injection or removal. On the 1st postoperative day, corneal biomechanical parameters as CH and CRF were significantly lowered in patients who underwent PPV with SF6 injection. In the rest of study groups following PPV procedures, CH and CRF did not change significantly on the 1st postoperative day. CCT values were significantly increased on 1st postoperative morning in eyes with air tamponade and oil tamponade in contrast to eyes with SF6 tamponade administered, where CCT was significantly decreased. Following oil tamponade removal, the CCT mean value did not change significantly on the 1st postoperative day. Although the results of measurements on the 1st day after surgery suggest differences in the parameters of the corneal biomechanics, the authors believe that this is the effect of poor reliability of the measurements, as evidenced by a significant reduction in the WS coefficient. WS coefficient (which is an indicator of the quality and reliability of the measurement) was significantly reduced in all groups of patients when measured on the first day after surgery.
In contrast, there were no significant changes of WS values on 7th postoperative day, when compared to the preoperative values in all study groups. This observation may suggest that the measurements of IOP and biomechanical properties of the cornea made on the 7th postoperative day were burdened with a smaller measurement error than those made on the 1st day after PPV.
Our results showed no statistically significant change in corneal parameters including CCT, CH and CRF as assessed on the 7th day after the PPV with air or SF6 tamponade. Similarly, CCT and CRF did not change significantly on the 7th day after silicone oil injection, however the mean value of CH decreased significantly in this group of patients. The lowering of the CH, in turn, can be related to a significant rise in IOPcc and GAT, observed in these patients on 7th postoperative morning. This relationship proved by our study in eyes with silicon oil tamponade was also previously described in some clinical studies over the effect of rise in IOP on CH in glaucomatous eyes 21,22 .
In patients who underwent PPV surgery with removal of silicone oil tamponade, the CCT, CH and CRF values were significantly reduced when measured on 7th day postoperatively in comparison to preoperative values. The question arises, whether the reduction was related to the surgery itself or it was a result of 6-months presence of intraocular silicon oil tamponade or both. Current studies suggest that in eyes with chronically elevated IOP, CCT is significantly reduced, what can be consistent with long-term increase in IOP related to silicon oil tamponade lasting 6-months in our patients 23 . The pathomechanism of CCT changes in patients with elevated IOP is not entirely clear 24 . It is suggested that CCT reduction in glaucomatous eyes may be the effect of remodeling of the corneal stroma under the increased IOP and pressure exerted on stromal collagen fibers 25 . This, in turn, leads to the stretching of the cornea and in consequence its thinning 25 . As it was mentioned above, remodeling of the corneal stroma in response to chronically elevated IOP in case of on silicon oil tamponade can be responsible for the decreased values of CH. According to our results, CRF values were elevated in eyes with silicon oil and have been significantly reduced after its removal. It is suggested that elevated CRF is a derivative of increased corneal stiffness through a transient tension in the stromal collagen fibers in response to an increase in IOP 26 . As we suppose, after the removal of silicone oil and reported significant decrease in IOPcc and GAT, the taut stromal fibers of the cornea relaxed leading to postoperative reduction of CRF.
Our results revealed a strong positive correlation between CH and CRF in all study groups pre and postoperatively, except patients who underwent silicon oil removal. In the other words, the corneal viscoelasticity described by CH correlated with stiffness and total resistance of the cornea described by CRF in patients referred to PPV, as well as in eyes that have undergone PPV with air, SF6 or silicon oil injection. This relationship between CH and CRF was not observed in patients prior, as well as on 7th day after silicon oil tamponade removal. What is more, there was a strong preoperative and postoperative correlation of CCT only with CRF but not with CH. This result is a strong argument that the silicon oil tamponade affects corneal viscoelasticity (CH), its stiffness (CRF) and thickness (CCT) in different ways, and this effect persists even after the removal of tamponade.
The major limitation of our study is the retrospective design. Further prospective clinical trials with longer follow up time and larger sample size are needed to expand the results of the study. Moreover, the duration time of silicone oil tamponade in the fourth study group was at least 6 months, which could have caused structural changes in the studied eyes. The morphological changes could have had an influence on the IOP, corneal biomechanics and CCT values measured, causing bias in the obtained results.

Conclusions
The PPV with intravitreal air or 25% SF6 injection does not significantly influence GAT and IOPcc, while PPV with silicone oil tamponade procedure causes significant rise in GAT and IOPcc values when measured on the 7th postoperative day. In turn, the removal of silicone oil tamponade causes a significant drop in both GAT and IOPcc. GAT and IOPcc did not differ significantly but correlated positively in pre-and postoperative assessments, which can prove that both measures are similarly reliable in patients after PPV procedure. The PPV with intravitreal air or 25% SF6 injection does not influence CCT and its biomechanical parameters including CH and CRF, when assessed on the 7th postoperative day. Similarly, PPV with silicone oil tamponade does not change the values of CCT and CRF, but it causes a significant decrease in CH. In turn, the PPV with removal of silicone oil tamponade causes significant decrease in CCT, CH and CRF values. Finally, there is a strong positive correlation between CH and CRF in patients who undergone PPV with air, SF6 or silicon oil injection. However, this www.nature.com/scientificreports/ relationship between CH and CRF was not observed in patients just after the removal of silicon oil tamponade. This result suggests that the chronic oil tamponade decreases corneal viscoelasticity (CH) and its stiffness (CRF) indeed but in a different way as evidenced by the lack of correlation between these parameters. The solution to this problem remains unclear and will become the subject of further research on the impact of intraocular surgical procedures on corneal biomechanics and precise IOP measurement.

Data availability
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.