Corneal Biomechanics as a Moderator of Central Corneal Thickness Changes After Phacoemulsification

1. Introduction

Cataract surgery is associated with early central corneal thickness (CCT) increase [1,2,3], endothelial cell loss [4,5,6], and transient alterations in corneal biomechanics [1,7,8,9]. Early postoperative corneal edema appears to reflect, at least in part, endothelial injury, as greater CCT increase has been associated with subsequent endothelial cell loss [10,11]. Both postoperative CCT increase and endothelial cell loss have been associated with several intraoperative parameters, such as effective phacoemulsification time, ultrasound exposure, cumulative dissipated energy (CDE), operative duration, and irrigation volume [4,12,13,14,15,16]. In contrast, the determinants of postoperative CH alterations are less well characterized. Kandarakis et al. [1] found that effective phacoemulsification time correlated with postoperative CCT increase but not with CH reduction, suggesting that the biomechanical response reflects broader effects of cataract surgery rather than ultrasound exposure alone. Moreover, Corvis ST studies have reported postoperative biomechanical alterations that were not fully accounted for by concurrent changes in pachymetry and intraocular pressure, further supporting only partial overlap between pachymetric and biomechanical responses [17,18].

The magnitude of postoperative corneal injury also varies according to individual ocular susceptibility. Age, endothelial reserve, cataract density, and anterior chamber depth (ACD), have been associated with endothelial loss or postoperative corneal edema [14,19]. Preoperative CH may represent an additional dimension of this susceptibility by providing biomechanical information complementary to conventional structural and biometric parameters.

Among intraoperative parameters, cumulative dissipated energy (CDE) is particularly relevant because it provides an integrated measure of ultrasound energy delivery during phacoemulsification, incorporating both the magnitude and duration of ultrasound application. In contrast, phacoemulsification time, operative duration, and irrigation volume capture only individual components of the surgical procedure and may not directly reflect the total ultrasonic energy applied. CDE has been widely used as a quantitative measure of ultrasound energy exposure during phacoemulsification [20,21] and has been associated with endothelial cell loss and early postoperative corneal edema [13,14,15].

However, the relationship between preoperative CH and the corneal response to CDE remains largely unexplored. We therefore hypothesized that the degree of postoperative corneal thickening associated with increasing CDE would vary according to the preoperative biomechanical state of the cornea, as reflected by CH. Clarifying this relationship may help define the potential value of CH as a preoperative indicator of individual susceptibility to phacoemulsification-related stress beyond structural and biometric parameters.

Given the apparent divergence between pachymetric and biomechanical responses after cataract surgery, further evaluation of their postoperative course and interrelationship is also needed.

To address these gaps, the primary aim of the present study was to determine whether preoperative CH modifies the association between CDE and early postoperative corneal thickening following uncomplicated phacoemulsification. Secondary objectives were to characterize postoperative changes in CH and CCT, assess their temporal recovery, and explore whether ocular biometric and clinical characteristics were associated with differences in corneal response.

2. Materials and Methods

2.1. Study Design and Participants

This prospective study included 89 eyes of 80 patients undergoing phacoemulsification cataract surgery and intraocular lens (IOL) implantation at Mureș County Hospital, Ophthalmology Clinic in Târgu Mureș, Romania between March-July 2026. Patients were eligible if complete preoperative and early postoperative measurements of CCT and CH, together with CDE, were available, and only eyes undergoing uncomplicated phacoemulsification cataract surgery were included. Eyes with pseudoexfoliation syndrome (PEX), glaucoma, or corneal guttata/ Fuchs endothelial corneal dystrophy (FECD) were retained for exploratory diagnosis-specific analyses. Exclusion criteria were ocular infection, central corneal opacities, keratoconus, previous corneal or refractive surgery, and current topical ocular or systemic corticosteroid treatment. Demographic data, corneal biomechanical parameters, and ocular biometric characteristics were collected for all participants.

2.2. Clinical Assessment and Outcome Measures

Preoperative evaluation included demographic data, central corneal thickness (CCT), corneal hysteresis (CH), corneal-compensated intraocular pressure (IOPcc), Goldmann-correlated intraocular pressure (IOPg), anterior chamber depth (ACD), and axial length (AL). CH was measured using the Ocular Response Analyzer (ORA; Reichert Technologies, Depew, NY, USA), and CCT using anterior pachymetry with spectral-domain optical coherence tomography (RTVue XR Avanti, Optovue Inc., Fremont, CA, USA). ACD and AL were measured by contact applanation ultrasonography using the OcuScan RxP Ophthalmic Ultrasound System (Alcon Laboratories, Inc., Fort Worth, TX, USA). All measurements were performed by the same experienced investigator using standardized central probe positioning and minimal corneal contact to minimize variation in probe alignment and contact pressure. The device-generated mean of 10 accepted measurements was used for analysis.

Pre-existing ocular conditions were documented from medical history and ophthalmic examination. Eyes were classified as controls or as having glaucoma, pseudoexfoliation syndrome (PEX), or corneal guttata; control eyes had none of these conditions.

CCT and CH were reassessed on postoperative day 1, and at one month. The primary outcome was the absolute early postoperative change in CCT (ΔCCT). Postoperative change in CH (ΔCH) and longitudinal changes in CCT and CH were evaluated as secondary outcomes.

2.3. Surgical Procedure

All patients underwent standard phacoemulsification cataract surgery using the CONSTELLATION Vision System (Alcon Laboratories, Inc., Fort Worth, TX, USA). All procedures were performed by a single experienced surgeon (KH) according to a standardized surgical protocol. A 2.2-mm clear-corneal main incision and two 1.2-mm side-port incisions were created. Dispersive and cohesive ophthalmic viscosurgical devices were used in a soft-shell technique to protect the corneal endothelium and anterior chamber structures. Phacoemulsification was performed using a vertical chop technique. CDE was automatically recorded by the phacoemulsification system and used as a measure of ultrasound energy delivered during surgery. At the end of the procedure, intracameral cefuroxime was administered for endophthalmitis prophylaxis.

2.4. Statistical Analysis

The primary analysis used multivariable linear regression with early postoperative ΔCCT as the dependent variable. Preoperative CH, CDE, baseline CCT, and age were entered as continuous predictors and mean-centered before model fitting. To test if the association between phacoemulsification energy and postoperative corneal thickening varied according to preoperative corneal biomechanics, a CH-CDE interaction term was included. The interaction was evaluated using its regression coefficient and post-estimation testing. Marginal predictions were subsequently estimated across representative preoperative CH values to illustrate how the association between CDE and predicted ΔCCT varied according to baseline CH.

In eyes with available biometric measurements, ACD was added to the multivariable model to assess its independent association with early postoperative ΔCCT after adjustment for CDE, preoperative CH, baseline CCT, and age. ACD was also evaluated as a continuous modifier of the CDE–ΔCCT association. An exploratory threshold-based analysis compared eyes with very shallow anterior chambers (ACD ≤2.5 mm) with those with ACD >2.5 mm, focusing on potential effect modification of the CDE–ΔCCT relationship at very low ACD. HC3 heteroskedasticity-robust standard errors were used because of the small subgroup size and potential influence of high-leverage observations.

Diagnostic subgroup analyses examined early postoperative corneal responses according to pre-existing ocular pathology. Eyes were classified as controls, glaucoma, PEX, or corneal guttata. Multivariable linear regression compared early postoperative ΔCCT with controls after adjustment for CDE, preoperative CH, baseline CCT, and age. CDE exposure across diagnostic groups was compared using the Kruskal–Wallis test. Longitudinal CH changes were evaluated relative to controls using mixed-effects models with diagnosis-by-time interaction terms. Given the small subgroup sizes, these analyses were considered exploratory.

Because CH is influenced by intraocular pressure and CH, IOPcc, and IOPg are derived from the same ORA measurement, sensitivity analyses were performed to assess whether the observed CH-CDE interaction was dependent on baseline IOP. Baseline IOPcc and IOPg were entered separately into the primary multivariable model including preoperative CH, CDE, the CH-CDE interaction, baseline CCT, and age. The two IOP parameters were not included simultaneously because of their strong intercorrelation and shared derivation from the ORA signal. In an additional exploratory model examining early postoperative CH change, baseline IOPcc was included together with CDE, baseline CH, baseline CCT, and age.

Longitudinal changes in CCT and CH were assessed using mixed-effects models. Pairwise comparisons between time points were derived from estimated marginal means and adjusted for multiple testing using the Bonferroni method. Model assumptions were assessed using residual diagnostics; multicollinearity, heteroskedasticity, and potentially influential observations were evaluated for the linear regression models. All tests were two-sided, and p<0.05 was considered statistically significant. Statistical analyses were performed using Stata, version 17 (StataCorp LLC, College Station, TX, USA).

2.5. Ethics Approval

The study was conducted in accordance with the Declaration of Helsinki and approved by the ethics committee of the hospital where the study took place (2374/18.03.2026). Written informed consent was obtained from all subjects before procedures.

3. Results

A total of 89 eyes were included in the analysis. Baseline demographic, corneal, and biometric characteristics are summarized in Table 1. Data are presented as mean ± standard deviation, median [interquartile range] or n (%), as appropriate.

3.1. Early Postoperative Corneal Response and CH–CDE Interaction

On postoperative day one, mean CCT was 687.1 ± 121.1 µm, corresponding to an absolute increase of 149.0 ± 112.2 µm and a relative increase of 27.6 ± 20.6%. Mean CH decreased to 9.28 ± 2.01 mmHg, with a mean change of −1.60 ± 1.98 mmHg from baseline.

A significant interaction between preoperative CH and CDE was observed in relation to early postoperative ΔCCT (Table 2). At mean CDE exposure, the model estimated a smaller postoperative CCT increase with higher CH, with the CH simple effect showing a trend toward statistical significance. The association between higher CH and lower predicted ΔCCT progressively attenuated as CDE increased (Figure 1).

The CH-CDE interaction remained unchanged after additional adjustment for baseline IOPcc or IOPg (β = 2.00, 95% CI [0.08, 3.92]; p = 0.041). Neither baseline IOPcc nor IOPg was independently associated with postoperative ΔCCT (p = 0.778 and p = 0.763, respectively).

3.2. Biometric and Diagnostic Subgroup Analyses

In the subset with available biometric measurements (n=67), neither ACD (β=4.2, 95% CI [−49.97, 58.52]; p=0.875) nor AL (β=−9.56 µm/mm, 95% CI [−26.27, 7.16]; p=0.257) was independently associated with early postoperative ΔCCT after adjustment for CDE, preoperative CH, baseline CCT, and age. ACD also did not modify the association between CDE and ΔCCT when analyzed continuously across the full biometric range (β=−2.35, 95% CI [−10.74, 6.03]; p=0.576). However, in the threshold-based analysis, shallow anterior chambers (ACD ≤2.5 mm; n=10) significantly modified the association between CDE and early postoperative ΔCCT (β=7.70, 95% CI [0.84, 14.56]; p=0.028). The estimated CDE–ΔCCT slope was 7.36 µm per CDE unit in eyes with ACD >2.5 mm and 15.06 µm per CDE unit in eyes with ACD ≤2.5 mm. CDE exposure was comparable between eyes with ACD ≤2.5 mm and those with deeper anterior chambers (p=0.94).

Early postoperative corneal responses varied according to diagnosis. Eyes with PEX or corneal guttata showed a tendency toward greater adjusted ΔCCT than controls (+45.4 µm; p=0.074), whereas glaucomatous eyes showed a tendency toward lower ΔCCT (−43.6 µm; p=0.083). CDE exposure did not differ significantly across diagnostic groups (p=0.154). In contrast, glaucomatous eyes exhibited a greater early postoperative CH decrease than controls (p=0.039), whereas the CH trajectory in eyes with PEX or corneal guttata did not differ significantly from controls (p=0.540).

3.3. Postoperative CH Response and CCT–CH Relationship

Higher CDE was weakly associated with a greater early postoperative reduction in CH (r=−0.222, p=0.036). After adjustment for baseline CCT and age, the effect remained small and showed only a trend toward statistical significance (β=−0.057 mmHg per CDE unit; p=0.054). In contrast, the magnitude of early postoperative CCT increase was not associated with CH reduction (r=−0.155, p=0.146) and remained unrelated to ΔCH after adjustment for CDE, baseline CH, baseline CCT, and age (p=0.786).

The association between CCT and CH was positive preoperatively (β=0.0094 mmHg/µm; p=0.047) but became significantly inverse on postoperative day 1 (β=−0.0027 mmHg/µm; p=0.034). At late follow-up, the association was again positive, with only a trend toward statistical significance (β=0.0122 mmHg/µm; p=0.085).

Follow-up measurements were available for 40 eyes. CCT was not significantly different from the preoperative value (MD=−14.63 µm, 95% CI [−48.45, 19.20]; p=0.902). CH showed substantial recovery from the early postoperative decrease but remained slightly lower than preoperatively (MD=−0.68 mmHg, 95% CI [−1.30, −0.06]; p=0.025).

4. Discussion

The main findings of the present study can be summarized as follows: (i) preoperative CH significantly modified the association between phacoemulsification energy and early postoperative corneal thickening, with higher CH being associated with a smaller predicted CCT increase at low-to-moderate CDE levels and this moderating effect progressively diminishing as CDE increased; (ii) although CDE was a strong determinant of postoperative CCT increase, its association with CH reduction was substantially weaker, while the magnitude of corneal swelling itself was not associated with CH decrease, suggesting that additional factors contribute to the early biomechanical response; and (iii) CCT returned toward preoperative values earlier than CH, indicating only partial overlap between pachymetric and biomechanical recovery. In secondary analyses, the increase in postoperative CCT per unit of CDE was approximately twice as large in eyes with shallow anterior chambers as in those with deeper chambers, suggesting greater sensitivity to phacoemulsification energy.

The interaction between preoperative CH and CDE indicates that the corneal response to phacoemulsification energy varies according to the preoperative biomechanical state of the cornea. At low-to-moderate CDE levels, higher CH was associated with a smaller predicted postoperative CCT increase, whereas this association progressively attenuated and became negligible at higher CDE levels. The interaction remained unchanged after adjustment for baseline CCT, age and baseline IOPcc or IOPg, supporting an independent association. Nevertheless, it should be interpreted as evidence of differential susceptibility rather than as a clinically meaningful threshold.

A related finding was that CDE was strongly associated with early postoperative CCT increase but only weakly with CH reduction, consistent with previous evidence of a partial dissociation between pachymetric and biomechanical responses to phacoemulsification [1]. To date, the relationship between intraoperative energy exposure and postoperative CH change remains poorly characterized. Our findings therefore add to limited evidence suggesting the involvement of additional determinants beyond ultrasound energy exposure and concurrent corneal edema.

Perioperative changes in ocular mechanical loading may represent one such influence. During phacoemulsification, irrigation and aspiration generate substantial transient variations in anterior chamber pressure, resulting in marked intraoperative IOP fluctuations [22,23,24]. Higher fluidic settings and intraoperative pressure have been associated with greater early postoperative corneal thickening, whereas lower target IOP with active-fluidics systems has been associated with smaller early CCT increases [25,26,27]. Although a direct effect of these pressure dynamics on postoperative CH has not been established, their contribution is plausible given the known dependence of corneal deformation and CH on mechanical loading and IOP [28,29,30].

Incision-related structural changes may also contribute to postoperative biomechanical alterations, although evidence regarding the influence of incision characteristics on the magnitude and recovery of these changes remains mixed [18,31,32]. However, because a standardized 2.2-mm clear-corneal incision was used in all eyes in the present cohort, incision size cannot explain the interindividual variability in CH change. Rather, the incision likely represents a common biomechanical perturbation, while differences in ocular loading and tissue hydration may contribute to variability in CH recovery [29,30,33].

The different temporal behavior of CCT and CH provides further evidence that pachymetric and biomechanical recovery only partially overlap. In the present study, CCT was no longer significantly different from its preoperative value at late follow-up, whereas CH, despite substantial recovery from postoperative day one, remained slightly but significantly reduced. Previous studies [3] have generally shown a transient reduction in CH after cataract surgery, although reported recovery ranges from normalization within one week [8] to approximately one month [9] while Corvis ST data suggest that altered biomechanical deformation may persist for up to three months even after accounting for IOP and pachymetric changes [18]. This difference is biologically plausible because CCT primarily reflects corneal thickness and hydration, whereas CH reflects the dynamic response of the cornea to loading and is only partly related to CCT [29,30,33]. Thus, pachymetric normalization should not necessarily be interpreted as complete biomechanical recovery. Persistent biomechanical alteration after pachymetric normalization may also be relevant when interpreting postoperative applanation-based IOP measurements, particularly in glaucoma patients [18,34].

ACD was not independently associated with early postoperative corneal thickening and did not significantly modify the effect of CDE when analyzed continuously. However, in eyes with shallow anterior chambers, the increase in postoperative CCT per unit of CDE was approximately twice that observed in deeper chambers, indicating greater susceptibility to phacoemulsification energy. Previous studies have reported greater endothelial injury in shallow anterior chambers, particularly in eyes with denser cataracts or under conditions of greater surgical stress, although findings have varied according to surgical technique and other operative factors [19,35,36]. Our findings extend these observations by indicating that ACD may influence the corneal response to a given amount of CDE rather than merely identifying susceptibility under high-energy conditions. The steeper CDE–CCT relationship in shallow chambers was evident across the observed CDE range, including at low-to-moderate levels, suggesting that increased susceptibility may also be relevant at lower energy exposures. This threshold-based finding should nevertheless be considered hypothesis-generating, as only 10 eyes had ACD ≤2.5 mm, and requires confirmation in larger cohorts.

Diagnosis-specific responses also supported a partial dissociation between pachymetric and biomechanical changes. Glaucoma, PEX, and corneal guttata/FECD have been associated with altered corneal biomechanics [37,38], although postoperative CH data remain limited [39,40]. In our cohort, glaucomatous eyes showed a greater early CH decrease despite a tendency toward less corneal thickening, whereas PEX or corneal guttata tended toward greater ΔCCT without a significantly different CH trajectory. These findings are consistent with previous reports of greater endothelial injury in PEX and increased postoperative edema susceptibility in corneal guttata [41,42]. Given the small subgroups, these results require confirmation in larger prospective studies.

From a clinical perspective, these findings support the potential value of CH as a complementary preoperative marker of individual corneal susceptibility to phacoemulsification-related stress. An integrated assessment incorporating biomechanical, structural, and biometric characteristics may allow a more nuanced characterization of corneal response and postoperative recovery.

Several limitations should be acknowledged. First, although the CH–CDE interaction reached statistical significance, the effect was modest and was identified in a relatively small cohort of 89 eyes; therefore, it should be interpreted cautiously and confirmed in larger independent samples. Second, late follow-up measurements were available in only 40 eyes, limiting the precision of conclusions regarding biomechanical recovery. Biometric measurements were available only in a subset of eyes, and the shallow-ACD subgroup was small. In addition, ACD was measured by contact applanation ultrasonography, which may introduce some underestimation through corneal indentation. However, all measurements were obtained by the same experienced investigator using a standardized technique to minimize measurement variability. CDE provides an integrated estimate of ultrasound energy exposure but does not capture other components of intraoperative mechanical stress, including pressure fluctuations and fluidic dynamics, which were not directly measured. Diagnosis-specific analyses involved small subgroups and should therefore be regarded as exploratory. Finally, only uncomplicated cataract procedures were included; although this improves internal consistency, the findings may not generalize to eyes experiencing substantial intraoperative endothelial trauma.