Myopia is the most common refractive error worldwide, and its prevalence continues to rise, making it a major public health concern (1). For adult myopic patients who are not suitable candidates for corneal laser surgery, posterior chamber phakic implantable collamer lens (ICL) implantation has become one of the mainstream refractive options because of its excellent visual quality, predictability, and long-term safety (2, 3).

Postoperative vault is a key determinant of the safety of ICL implantation (4). Insufficient vault may lead to contact between the ICL and the anterior capsule of the crystalline lens, increasing the risk of anterior subcapsular opacification and cataract, whereas excessive vault may push the iris forward, impair aqueous outflow, and trigger intraocular pressure elevation or even pupillary block (5, 6). The central hole in the V4c design has been shown to relatively reduce some of these risks, but it does not eliminate these risks entirely (7, 8). Furthermore, the inherent optical design of the V4c lens, characterized by a flat anterior surface and increased edge thickness, particularly in high myopic corrections, can elevate the risks of angle crowding, pigment dispersion, and potential endothelial touch (7, 9).

Vault is largely determined by the match between ICL size and the patient’s intraocular anatomy, among which the sulcus-to-sulcus (STS) distance is a critical parameter (9). In current clinical practice, ICL sizing is still mainly guided by the STAAR company nomogram based on external measurements such as horizontal corneal diameter (white-to-white, WTW) and anterior chamber depth (ACD) (9, 10). However, previous studies have shown that the correlation between WTW and the true STS distance is limited, which may contribute to suboptimal postoperative vault (11, 12).

Ultrasound biomicroscopy (UBM) can clearly visualize the ciliary body and ciliary sulcus and is regarded as the gold standard for measuring STS (13). Several studies have reported that incorporating UBM-derived STS into ICL sizing can improve the predictability of postoperative vault (9, 14). Notably, the ciliary sulcus is not perfectly circular in most eyes, but tends to be vertically elliptical, with the vertical STS (vSTS) exceeding the horizontal STS (hSTS) (15). This anatomical asymmetry may vary with refractive status and directly influence ICL stability and postoperative vault (9, 15). Nevertheless, most existing studies have been limited to a single refractive error range or have focused only on hSTS, and systematic data on the distribution and differences of hSTS and vSTS across different myopia severities remain scarce (7, 9).

In addition, although UBM provides STS measurements that more closely reflect the true intraocular anatomy, to effectively integrate UBM-derived STS and other anterior segment parameters into preoperative planning and construct quantitative, accurate vault prediction models that offer intuitive decision support for ICL sizing still requires in-depth research. Therefore, this study aimed to comprehensively investigate the distribution and correlations of hSTS, vSTS, and their differences in myopic eyes with varying refractive errors, and to develop and preliminarily evaluate a regression model for predicting postoperative ICL vault based on key anterior segment parameters, with particular emphasis on STS, in order to provide new strategies for individualized ICL sizing and enhanced surgical safety.

ICL postoperative vault is a key determinant of safety after ICL implantation. Its magnitude depends largely on the match between ICL size and the patient’s intraocular anatomy and is shaped by a complex interaction of multiple factors (9, 16). Achieving precise preoperative selection, therefore, requires detailed anterior segment information that corresponds to the actual attachment site of the ICL footplates (5, 14). With respect to key parameters, standard clinical AS-OCT can provide several useful parameters, such as corneal curvature and ACD, but it cannot consistently and reliably visualize structures posterior to the iris for ciliary sulcus assessment in a routine clinical setting (12, 17). In contrast, UBM affords direct visualization of the ciliary body and sulcus and enables precise measurement of STS distances, providing a more dependable basis for evaluating ICL sizing and positioning (14). However, although a vertically elliptical ciliary sulcus configuration has been reported previously (18), the refractive-status–dependent changes in STS asymmetry have not been systematically characterized, which, to some extent, has limited the wider adoption of STS-based approaches for ICL sizing and vault prediction. Ultimately, the safety and efficacy of ICL surgery hinge on achieving an optimal biomechanical match between the lens and the posterior chamber space (9). This fundamental requirement underscores the enduring necessity for precise assessment of the ciliary sulcus anatomy, regardless of evolving ICL lens design (19, 20).

Our study employed SE for group stratification, as it is the principal clinical determinant of myopia severity in refractive surgery (21, 22). Although AL is a more direct geometric biometric, its use in clinical grouping lacks standardized thresholds (23–25). In the ciliary sulcus morphology cohort, UBM measurements were performed using a standardized protocol recommended by expert consensus in a relatively large sample of myopic eyes, and STS diameters were analyzed in fine 0.50-mm intervals. This approach systematically demonstrated that a vertically elliptical sulcus configuration (vSTS > hSTS) is ubiquitous in myopic eyes, consistent with previous studies in pseudophakic eyes (18), and further showed that this asymmetry becomes significantly more pronounced with increasing myopic SE. Accordingly, the diameter ratio hSTS/vSTS was further incorporated into the analysis, and the proportion of eyes with hSTS/vSTS < 1 was found to reach 100% in the extreme myopia group, providing morphologic confirmation that the degree of vertical ellipticity escalates as myopia worsens. These findings are consistent with previous reports and quantitatively supplement the description of this progressive pattern (1). In addition, the results from this morphology cohort suggest that the influence of SE on STS may involve a “dual mechanism.” First, a “global enlargement” effect was identified, in which both hSTS and vSTS increased as SE became more myopic, consistent with the axial length–based findings reported by Wang et al. (26) Second, a “progressive asymmetry” was observed, characterized by a greater vSTS-hSTS difference and a higher proportion of eyes with vSTS-hSTS ≥ 0.5 mm. Taken together, these observations indicate that the ciliary sulcus undergoes increasingly marked vertical elliptical remodeling as myopia deepens. Consequently, preoperative ICL planning should account for the differential impact of refractive status on horizontal and vertical STS, which is important for selecting lens size, preventing postoperative rotation, and maintaining a stable vault. These morphologic findings not only complement previous work on the spatial structure of the STS but also provide a clear anatomical basis for understanding the mechanical behavior of the ICL and the accompanying changes in vault after implantation.

Building on these morphologic findings, an independent vault-modeling cohort was used to develop a predictive model in which preoperative hSTS and ICL size were entered into a multiple linear regression with 1-month postoperative vault as the dependent variable. The model demonstrated good performance in a single-center dataset, with a simple structure, easily obtainable input parameters, and intuitively interpretable output. Compared with more complex machine-learning approaches that require large training datasets, this conventional regression model is more convenient to implement at the current sample size while still providing stable and reliable predictions, thereby enhancing its potential for clinical translation and wider application.

In clinical practice, ICL size is still most commonly selected on the basis of WTW and ACD (7, 27). However, several studies have shown that the correlation between WTW and STS is limited, and acceptable predictive accuracy is achieved only when the difference between the two lies within a narrow range; overreliance on WTW may therefore lead to inappropriate ICL sizing and abnormal postoperative vault (28). At the same time, previous reports have consistently suggested that ACD is closely associated with postoperative vault (7, 9, 10). In this study, however, once hSTS was included in the stepwise regression model, the independent associations of WTW and ACD with vault were no longer significant. This finding indicates that, within an STS-based sizing strategy, hSTS more directly reflects the key anatomical determinants of postoperative vault, whereas conventional anterior segment parameters derived from corneal and anterior chamber geometry make a relatively attenuated independent contribution.

In addition to WTW and ACD, the vault-modeling cohort also evaluated the contribution of other preoperative parameters, including corneal curvature and AL. Univariate analyses showed that K1 and K2 were moderately negatively correlated with postoperative vault, and AL was also associated with vault, suggesting that corneal shape and AL may participate in vault formation by altering anterior segment configuration. These observations are broadly consistent with previous reports indicating that corneal curvature and AL may influence vault primarily through their impact on ACD (23). However, in the multivariable regression analysis, the effect sizes of these parameters were relatively small, and their explanatory power for vault was clearly inferior to that of hSTS and ICL size; therefore, they were not retained in the final model. Likewise, vSTS was not included in the regression equation, which may reflect the fact that, in current clinical practice, ICLs are predominantly implanted along or near the horizontal meridian, and postoperative vault is thus mainly constrained by hSTS (9, 15). Future studies should investigate different implantation axes and rotational states to comprehensively characterize the relationships among hSTS, vSTS, vault, and ICL stability, with the aim of further refining STS-based individualized sizing and axis-planning strategies.

This study has several limitations. First, it was a single-center retrospective analysis conducted exclusively with the ICL V4c model, and both cohorts had relatively modest sample sizes; as a result, the external validity of the findings is limited, and they should be interpreted with caution until confirmed in larger, multicenter prospective studies. In the ciliary sulcus morphology cohort, refractive status was classified primarily on the basis of SE; AL and corneal curvature were not incorporated into the grouping scheme, nor was a distinction made between spherical and cylindrical components. Moreover, measurements were restricted to horizontal and vertical meridional STS and did not cover other STS orientations and adjacent structures such as the ciliary processes. So the depiction of the spatial structure of the STS still needs to be supplemented. In the vault-modeling cohort, the consistency of UBM measurements has not been systematically verified, and potential inter-device conformity among different UBM systems from various manufacturers has not been evaluated. In addition, the set of anterior segment parameters entered into the model remained relatively limited, without inclusion of finer anatomical indices such as iris thickness and ciliary process morphology. Furthermore, parameters describing the anterior curvature of the crystalline lens, such as crystalline lens rise (CLR), which may refine vault prediction (29), were not included in our analysis and represent an important direction for model optimization. Although ACD no longer showed an independent association with vault once hSTS was introduced into the multivariable model, the present data are insufficient to conclude that its role can be completely replaced by STS. Future multicenter studies with larger samples and a more comprehensive panel of anterior segment parameters, together with external validation, are needed to improve further the accuracy and robustness of STS-based vault prediction models.

This study confirmed that the ciliary sulcus in myopic eyes is predominantly vertically elliptical, and that the vSTS-hSTS difference increases with higher myopic SE, suggesting the need to routinely assess both horizontal and vertical STS—rather than relying solely on WTW or a single horizontal diameter—when planning ICL surgery in cases of high and extreme myopia. Based on this understanding, a simple vault prediction model incorporating hSTS and ICL size was constructed using readily obtainable parameters, and it may serve as a useful complement to existing sizing nomograms, providing a practical reference for individualized ICL selection and vault management in highly myopic patients.