With the advance in wavefront technology, customized selection of aspheric IOLs to compensate individual corneal aberrations has been a trend in the current era of refractive cataract surgery. The aspheric IOLs are designed to compensate the positive corneal SA and thus reduce the ocular SA. The availability of various aspheric IOL designs has enabled the customized selection to precisely achieve a postoperative target ocular SA of zero, which has been considered to be the optimal SA value for producing the best optical quality and visual performance by some researchers [20, 21]. However, there are few controlled studies on optical quality of customized selection with a target SA of zero and the conclusions are inconsistent [13, 14]. The present study aimed to comprehensively compare the postoperative optical quality using a combination of Hartmann-shack wavefront aberrometer and double-pass instrument between customized and random selections of aspheric IOLs, which has been rarely applied in previous studies regarding the customized selection with a target ocular SA of zero.
In this study, no significant differences in postoperative UCVA and BCVA were observed between the two groups, which were consistent with the findings of previous controlled studies related to customized selection [13, 14, 22, 23]. The finding of the mean corneal SA at 6mm PD was comparable to that in the study by Beiko et al.  (0.28μm vs 0.27μm), and the unchanged corneal SA after surgery indicated that a bias resulting from the surgical impacts on corneal SA could be avoided. With a 6mm PD, t-HOA, SA, coma and trefoil were all significantly lower in the customized group, indicating that customized selection reduced the ocular aberrations more effectively than random selection. Moreover, in the customized group, the mean postoperative ocular SA was 0.00μm, which indeed met the goal of this study, that was, rendering the postoperative ocular SA to be closest to zero. This result was also consistent with the previous studies [11, 12]. In the control group, the postoperative ocular SA was 0.12±0.18μm, indicating that almost half of the corneal SA remained by using a random selection, which was similar to the result in the study using the same selection strategy by Jia et al.  (a residual SA of 0.152±0.151μm). As is well-known, the coma and trefoil aberrations are mainly related to IOL tilt, decentration and surgical incisions, and the IOL tilt and decentration differ among different types of IOLs . Although the same incisional procedure was used and comparable tilt and decentration were presented in both groups, results demonstrated that coma and trefoil showed significant difference between the two groups when the pupil increased. The only controlled study on customized selection that reported coma and trefoil aberrations was by Nochez et al. , which compared a customized group targeting SA to +0.1μm with a control group receiving zero SA IOLs only. Results showed similar coma and trefoil in both groups. However, as mentioned, the study design and the IOLs used by Nochez et al. were quite different from the present study, which might lead to the inconsistent results. The unexpected results in this study might be due to the difference of IOL selection strategies in the two groups, which led to imbalance in IOL allocation and thus SA distribution of IOLs between groups. IOLs with various SA values could induce various degree of coma, trefoil and thus t-HOA, even with less noticeable tilt or decentration [16, 25]. However, we had a good reason to apply a random selection strategy as a control, which was the strategy we have been using in actual clinical practice in the past. Overall, the results regarding wavefront aberrations in this study suggested that customized selection of aspheric IOLs based on corneal SA is feasible to achieve a target ocular SA of zero, and a random selection lead to a positive residual SA of 0.12μm. Customized selection demonstrated a better optical quality in terms of wavefront aberrations at 6mm PD.
Although lower aberrations demonstrated in the customized group, similar values for all OQAS parameters were shown in both groups in the present study. The results for OVs that represent contrast visual acuity were consistent with Al-Sayyari et al.  who found no significant difference in contrast sensitivity between the customized group and the control group. The study by Nochez et al.  was the only study applying OQAS system to assess optical quality in customized selection, in which they found significantly better MTF cut-off, OV 100%, OV 20% and OV 9% in the customized group than in the control group. However, their study design was quite different from the present study. They set up the target SA to +0.1μm in the customized group and only two SA values (0μm and -0.18μm) of aspheric IOLs were available for selection. The control group received zero SA IOLs only. Eventually, they achieved a mean postoperative ocular SA of 0.085μm in the customized group and 0.261μm in the control group, the difference in the mean postoperative ocular SA between groups was 0.176μm in their study, which was much higher than that in the present study (0.12μm). This might be the main reason for the inconsistent results between the two studies. As was demonstrated in one of our previous studies, OQAS system could reveal inconsistent outcomes with aberrometers. In that study, we concluded that the optical quality might be overestimated when only aberrations were considered, thus combining the effect of ocular scattering with aberrations by an OQAS system would result in a more accurate assessment of optical quality . Both aberrations and intraocular scatter are known to decrease the quality of the retinal image and therefore the visual performance. Previous evidence suggested that OSI had no significant correlation with SA in pseudophakic eyes (r=0.133, P=0.525), whereas a significantly positive correlation with the total ocular aberrations (r=0.451, P=0.024) . Although the P value showed a statistically significance, the correlation coefficient was not large enough to suggest a clinical significance. This may explain why lower aberrations achieved by the customized selection did not lead to an OSI reduction compared to the random selection in the present study. Another previous study investigated the impacts of SA and intraocular scatter on contrast sensitivity, the results demonstrated that contrast sensitivity was reduced less by scatter when SA was present as compared with the cases without SA. They concluded that combined presence of positive SA and scatter could be a mild protective compensatory mechanism reducing the impact that the two factors may have on contrast sensitivity separately . This viewpoint may explain why the overall optical quality interpreted by MTF cut-off, SR and OVs in the control group, though had a larger positive SA, were not worse than the customized group in this study.
In addition, there is currently controversy over the optimal target ocular SA values for an aspheric IOL: 0μm or +0.10μm. Some researchers argued that IOL implantation aiming to reach a target ocular SA to zero could lead to the best optical quality and visual performance [12, 20, 21, 28, 29]. Nevertheless, some researchers argued the same for creating a slightly positive residual ocular SA [22, 23, 30, 31]. Several controlled studies confirmed that both currently argued modes of target ocular SA provided desirable optical quality and visual performance. Specifically, zero SA was more beneficial in mesopic condition while slightly positive SA was more beneficial for near vision [14, 29]. One of the reasons why the overall optical quality of the customized selection did not differ from the random selection might also be that the mean residual ocular SA by a random selection in the present study (+0.12μm) was fairly comparable to +0.1μm, which was argued by previous studies to provide a comparable optical quality to that of a zero target ocular SA.
Moreover, as suggested by several previous studies: 1) although SA can be strongly reduced by aspheric IOLs, t-HOA was only slightly reduced. This may be a reason for the unclear results in studies assessing the potential benefit to visual performance of aspheric IOLs ; 2) the optimal ocular SA producing the best image quality varied widely between eyes, which was highly associated with the amount of defocus. The amount of optimal ocular SA could be predicted based on other HOAs of the cornea as well. Selection of an aspherical IOL should be customized based on the full spectrum of corneal HOAs, not on SA alone ; 3) a study using adaptive optics vision simulator demonstrated that pseudophakic eyes experienced peak contrast sensitivity performance with varying levels of SA when their natural aberrations were present. However, average contrast performance peaked at SA of zero when all HOAs were corrected . These studies indicated that interactions between SA with other HOAs and lower-order aberrations are complicated and not well understood. Customized selection of aspheric IOL implantation based on corneal SA alone may not be an appropriate approach to achieve optimal visual outcomes. Conclusions of these studies may also explain why there was no significant difference in the overall optical quality between customized and random selections of aspheric IOLs in the present study.
Compared with the previous controlled studies, the strengths of the present study were a larger sample size, a more specific customized selection with narrower SA ranges, as well as a more comprehensive and objective assessment for optical quality. There are also some limitations: 1) the focus of the customized design was only on the optimization of postoperative SA; 2) only one target ocular SA value (0μm) was set up for the customized selection; 3) subjective assessment for visual satisfaction was not available. Further studies related to customized selection of aspheric IOLs should pay more attention to the full spectrum of preexisting corneal HOAs but not only SA . Further attentions should also be paid to investigate how tilt, decentration and depth of focus impact on the optimization of optical quality in customized selection. It is also necessary to explore the differences in postoperative optical quality by OQAS among customized selection with different target ocular SA values. In addition, visual satisfaction questionnaire should be applied to complement with the objective optical quality assessments.