In recent years, the influence of posterior corneal astigmatism has caused great concern in the clinical practice with the access to the measurement of posterior corneal astigmatism. Optical coherence tomography (OCT) and Scheimpflug anterior segment analysis systems can directly measure the posterior cornea surface, which is more accurate than corneal topography systems [20,21]. The reliability and repeatability of Pentacam have been clinically proven for corneal astigmatism measurements [22,23].
It has been reported that the posterior cornea is a minus lens, which is WTR astigmatism [24]. With the growth of age, the meridian of posterior corneal astigmatism remains stable, while the anterior corneal astigmatism changes from WTR to ATR. If the fixed ratio is used based on the anterior corneal astigmatism, an overcorrection of 0.5~0.6 D is caused in patients with WTR astigmatism, while an undercorrection of 0.2~0.3 D is caused in patients with ATR astigmatism [10, 13].
However, it still remains controversial whether we should treat the posterior corneal astigmatism as a fixed, minus value or dynamic parameter that changes after cataract surgery. There is a debate about whether a cataract surgical incision can cause significant P-SIA. According to the results of Nemeth et al. [15], P-SIA is non-negligible with an average of 0.31 D, and about 25% of patients have a P-SIA of 0.5 D or more. In addition, it is more likely to cause lager P-SIA among patients with toric IOL implantation. Cheng et al. [14] have also found that ignoring the P-SIA may cause calculation errors of SIA. However, some studies have concluded that the effect of the incision on the corneal posterior surface is negligible. For example, Klijn et al. [16] have found that the average P-SIA is 0.1 D. Kim et al. [25] have found that P-SIA is not uniform with an average of 0.20 ± 0.17 D.
Previous studies on P-SIA [15] have predicted that P-SIA may have an important impact on the patients who are suitable for toric IOL implantation. The biomechanical effect of the incision on patients with larger corneal astigmatism may be different from those with smaller corneal astigmatism, and the accurate estimation of SIA is especially important for toric IOL implantation. Therefore, patients who are suitable for toric IOL implantation were set as the research objects in the current study. In this study, the preoperative anterior corneal astigmatism was 1.59±0.69 D. Similar to previous studies [10], the posterior corneal astigmatism was 0.28±0.20 D, which was dominated by WTR (62.7%).
The method described by Holladay-Cravy-Koch, which is different from simple vector calculation and has been recognized as a more accurate calculation method, was used to calculate P-SIA in this study. We found that 26.7% of patients had a P-SIA greater than 0.5 D with an average of 0.34±0.20 D, suggesting that the effect of the incision on the posterior corneal astigmatism was significant and might affect the postoperative total corneal astigmatism. Besides, P-SIA has a large individual difference, possibly due to the difference in corneal rigidity, thickness, biomechanical condition and healing ability [26,27]. The P-SIA obtained in this study was not significantly different from the P-SIA obtained by Nemeth et al. [15] (P = 0.64), which is larger than the P-SIA obtained by Kim et al. [25] (P = 0.00). This discrepancy might be attributed to the different locations of incision. This study, as well as Nemeth’s study[15], used the steepest meridian incision, while the temporal side incision is used in Kim’s study [25], suggesting that the location of the incision might have an effect on the results.
In order to explore the influencing factors of P-SIA, P-SIA was divided into different groups based on the astigmatism axis on the posterior corneal astigmatism. The results showed that there was no significant difference among the three groups, indicating that P-SIA was not associated with the astigmatism axis of the posterior cornea. The Pearson’s correlation analysis showed that P-SIA had a significant positive correlation with anterior and posterior corneal astigmatism, suggesting that patients with large preoperative astigmatism on the corneal anterior or posterior surface might cause larger P-SIA.
This study indicated that cataract incisions caused a significantly reduced anterior corneal astigmatism but an increase of the corneal posterior astigmatism, and the P-SIA did not show reduction trend in patients with closer axis of anterior and posterior corneal astigmatism. This finding could probably be attributed that the posterior corneal astigmatism was smaller than anterior corneal astigmatism, and the biomechanical characteristics of posterior corneal were different from the anterior cornea, leading to the difficulty in releasing the posterior corneal astigmatism. We speculated that the shape of the incision, corneal thickness, rigidity, especially the distance from the end of the incision to the center of the cornea, might be related to the P-SIA[28–29]. For example, we inferred that with the shorter distance between the end of the incision and the center of the cornea, the P-SIA might be greater.
There are some advantages and significances in this study. (1) In this retrospective study, the inclusion and exclusion criteria were severely restricted to control the information bias. (2) This study considered relatively comprehensive factors that might be related to P-SIA, and we concluded that patients with larger corneal anterior or posterior astigmatism might have larger P-SIA. (3) As the accurate prediction of P-SIA of toric IOL is demanding, this study focused on patients who were suitable for toric IOL implantation and limited the corneal astigmatism of patients.
However, there are still some limitations in this study as follows. (1) The follow-up period of this study was relatively short. Although previous study has indicated that there is no significant difference of SIA between 1 month and 6 months after surgery [29], extending the follow-up time can avoid the influence of the difference in individual healing ability on the experimental results to a certain extent. (2) Although we only included the measurements with the QS reading “OK,” there might still be differences between measurement and re-measurement of the Pentacam result. (3) Some important factors, such as the distance from the end of the incision to the center of the cornea, corneal thickness and so on, were not investigated in this study due to the incomplete data. These limitations suggest that it is necessary to design a prospective, long-term study with a control group to verify the conclusion of this study.