A novel analysis of Scheimpflug total corneal refractive power following corneal cross-linking in mild to moderate keratoconus

INTRODUCTION

Keratoconus is a progressive, mostly bilateral corneal degenerative disorder characterized by apical thinning and local protrusion that can result in irregular astigmatism or even corneal scarring with impaired vision

. There exists no treatment therapy to stabilize or slow the progression until the introduction of corneal cross-linking (CXL), which was initially proposed by Wollensak

for treatment of progressive keratoconus by increasing the intrinsic corneal biomechanical stability in 2003. Currently, it has been accepted as the major treatment modality for progressive keratoconus since plenty of studies in the literature have demonstrated its short- and long-term safety and efficacy

.

Meanwhile, there were several prominent challenges in CXL, including no consistent or clear definition of ectasia progression

, tremendous variability in nomenclature and raw data reported

, and miscellaneous corneal parameters(topographic, tomographic, pachymetric, biomechanical,

.)in detection of CXL effectiveness. Usually, the maximum K value (Kmax) is the commonly used keratometric parameter to indicate the efficiency of CXL in previous studies

.Nonetheless, series of studies have proposed that the Kmax is a problematic index for the follow-up of keratoconus in terms of repeatability

. The Kmax, which neglects the contribution of the posterior corneal surface to progression, assumes the natural anterior-to-posterior curvature ratio and subsequently the arbitrary keratometric index (1.3375), and intends to evaluate the severity of corneal protrusion solely on the central area or a single point other than the keratectasia area, has been doubted as a reliable parameter for both progression detection and crosslinking efficacy evaluation

.

Furthermore, our study evaluated the performance of the third version of Kmax (zone value)-Km

, which may more globally represent the ectatic region than Kmax as a single point parameter. Interestingly, Km

3 mm in Group 2 exhibited a slight improvement early at 3mo while Ksteep and Kmax showed reduction until 6mo postoperatively. Previously,Lytle

reported that the Km

3 mm not Kmax showed statistically significant improvement from baseline at 3mo and proposed that the 3 mm zonal Km may allow for earlier detection of the efficacy of CXL than the use of a point Kmax value alone. Meanwhile, a recent study

disclosed that the keratectasia area showed in corneal topography was more sensitive than Kmax/Ksteep in describing the early stage morphological changes in keratoconic eyes. Taking all these works together, it seems to be speculated that a zonal Km centered on Kmax taking in account a larger portion of the ectatic region rather than a single point value may more accurately depict changes in the keratoconus remission or deterioration following CXL treatment. Nonetheless, such a mild improvement (0.07 D) presented in the current study failed to achieve a statistical significance and limited its clinical application. Besides, it is unclear why Km

3 mm other than 1 mm or 2 mm could exhibit such an earlier mild improvement trend. One possible explanation could be that the peripheral ectatic area may show more prominent topographic changes compared to the more central ectatic area. A larger sample study would be warranted to explore the possible mechanism and to validate if zonal Km centered over the cone could serve as an earlier indicator in detecting the effect of CXL on keratoconus stabilization and improvement.

SUBJECTS AND METHODS

The study was performed in accordance with the tenets of the Declaration of Helsinki and approved by the Ethics Committee and Institutional Review Board of Hankou Aier Eye Hospital. Progressive keratoconus patients who had received CXL treatment between August 2017 and September 2018 in Hankou Aier Eye Hospital were retrospectively enrolled into the current study. All participants were provided with written informed consent before participation.

Finally, as a second outcome, the changes of keratometric parameters in keratoconic eyes after CXL treatment displayed two distinct trends during 12mo follow-up. Majority of the keratoconic eyes displayed an initial deterioration at 1mo with subsequent improvement up to 12mo postoperatively, which corroborates with previous studies

. It is speculated that the apparent initial keratoconus progression at 1mo could be attributable to the epithelial debridement and regrowth

. In 2009, Reinstein

described that the corneal epithelium by Artemis very high-frequency digital ultrasound system in keratoconic eyes demonstrated an epithelial doughnut pattern of a localized thinning over the cone surrounded by an annulus of epithelial thickening, which could either partially or totally mask the stromal surface cone. In accordance with the findings of Reinstein

, previous studies observed that epithelial debridement increased the magnitude of anterior corneal keratometry in keratoconic eyes

. In addition, the regrowth of an epithelial layer of more uniform thickness after CXL procedure could paradoxically result in apparent steepening of the keratometric parameters until the halt effect by increased biomechanical stability prevails over the epithelial remodeling effect

. There existed a second change trend in the present study that minority keratoconic eyes exhibited a flattening initially at 1mo, achieved the largest flattening effect at 3mo,and subsequently reduced gradually till 12mo, which is in alignment with previous findings of Asri

Previously,Caporossi

conducted a prospective study in evaluation of stability and functional response after CXL in pediatric patients with keratoconus, in which patients in the thinner group (TCT＜450 μm) showed a significantly faster functional recovery than the thicker group (TCT＞450 μm) at the 3-month follow-up. In the current study, baseline characteristics analysis suggested that higher manifest refraction spherical equivalent, worse visual acuity may take response for an initial improvement process following CXL treatment. Meanwhile,the keratometric readings and pachymetric parameters in Group 1 reflected a slightly serve degree but without statistically significance. It is speculated that the severity degree of disease, the CXL procedure, the respond to the CXL treatment, the corneal epithelium remodel process may play an important role in the change trend and recovery speed.Nonetheless, a prospective study may further disclose the possible explanations for the above phenomenon.

In Group 2, all keratometric parameters exhibited steepening at 1mo, subsequently relieved gradually and achieved relatively stable improvement at 12mo except for Km

1 mm(mean change -0.69 D,

=0.19). Kmax exhibited the most improvement (0.92 D) at 12mo, followed by TCRP

2 mm (0.91 D) and TCRP

1 mm (0.87 D; Table 4).Besides, only TCRP

within diameters of 1 mm to 3 mm and Km

3.0 mm became flatter than the preoperative level early at 3mo postoperatively, while other keratometric parameters showed flatter than the preoperative level until 6mo postoperatively. TCRP

3 mm showed the most improvement(0.19 D), followed by TCRP

2 mm (0.15 D; Figure 2).

CXL procedures were performed by one experienced surgeon (Chen D) using the epithelium-off CXL as described in previous study

(Table 1).

1.1 資料來源 選擇2015-2016年于江西省贛州市人民醫院進行產前檢查與分娩的孕產婦600例為研究對象，年齡21～38歲，平均年齡(28.36±2.83)歲；孕周37～42周，平均孕周(40.08±1.13)周。納入標準：①年齡≥20歲；②初產婦、單胎；③婚姻關系正常；④依從性較佳，可有效完成隨訪調查；⑤自愿接受調查，簽署知情同意書，本研究經醫院倫理委員會批準。排除標準：①智力障礙、患有精神疾病，認知功能不全；②有糖尿病史，心、肝、腎等疾病史；③高危妊娠；④合并妊娠期高血壓疾病；⑤終止妊娠者。

All the data were analyzed using SPSS software version 25 (IBM Corp., USA) for Windows.Data were expressed as mean±standard deviation (SD) if fulfilled normality test using the Kolmogorov-Smirnov test.Otherwise, the results were expressed as the median (25

and 75

percentile). One-way analysis of variance (ANOVA) for repeated measures with Bonferroni multiple comparisons was used to assess the time course of changes following CXL treatment in series of keratometric parameters for normally distributed data. For two-sample comparison, statistical analysis was conducted by the Mann-Whitney

test for nonparametric data (refraction, UDVA, and CDVA) and by the unpaired

-test for parametric data (Kmax, Ksteep,

.). A level of

＜0.05 was chosen as the criterion for significance.

RESULTS

雞皮刺螨感染引起的鵝體表寄生蟲病要選擇高效驅蟲藥物進行針對性治療。本次治療選擇使用伊維菌素，使用劑量為2 mg/kg體重，連續使用5 d。同時，將養殖舍內的墊料全部清理，堆積發酵，殺滅墊料中存在的成蟲。選擇使用0.005%的溴氰菊酯或精制敵百蟲外用藥物，配成2%的溶液對養殖舍墻角、縫隙、飲水器、水管等處進行噴灑消毒，一定要噴灑徹底，上述藥物交替使用，1周噴灑2次，連續使用2周。通過采用上述綜合處理手段處理2周后，該養殖場有75只鵝死亡，剩余患病鵝恢復健康，病情得到切實有效的控制。

Kmax showed the most improvement at 12mo postoperatively(1.00 D), followed by TCRP

1 mm/2 mm (0.96 D and 0.92 D, respectively). All keratometric parameters exhibitedslight increase from preoperative to 1mo postoperative, then decreased gradually to 12mo (Table 3). Most keratometric parameters displayed significant reduction as early as 6mo postoperatively, while only TCRP

3 mm and TCRP

1/2/3 mm achieved significant improvement at 3mo postoperatively (

＜0.05). TCRP

1 mm obtained the most improvement (0.57 D), followed by TCRP

2 mm (0.54 D)and TCRP

3 mm (0.43 D).

The percentage of change of Kmax at 12mo less than -0.50 D,within ±0.50 D and more than 0.50 D were 60% (24/40), 30%(12/40), and 10% (4/40), respectively. While there were 57.5%(23/40), 35% (14/40), and 7.5% (3/40) change of TCRP

2mm at 12mo less than -0.50 D, within ±0.50 D, and more than 0.50 D, respectively.

Unlike the overall, all keratometric parameters in Group 1 exhibited flattening initially at 1mo, achieved the most flattening effect at 3mo, and subsequently reduced gradually till 12mo. At 12mo,all keratometric parameters showed significant improvement compared to baseline except for Ksteep, in which Kmax displayed the most improvement (1.13 D), followed by TCRP

1 mm (1.10 D) and 2 mm (0.95 D; Figure 2).Most keratometric parameters displayed significant reduction at 3mo, while only Kmax, Km

2 mm, and TCRP

2 mm achieved significant improvement early at 1mo postoperatively (

＜0.05; Table 4).

The severity of keratoconus was graded according to the Amsler-Krumeich classification system which depends on corneal power, corneal thickness, astigmatism, and corneal transparency. Progression was defined as one or more of the following changes over a period of 12mo: an increase of at least 1 diopter (D) in the Kmax measurement, an increase of at least 1.0 D in manifest cylinder

. Exclusion criteria includedthinnest corneal thickness (TCT) less than 400 μm, central or paracentral opacities, any previous ocular surgery, a history of chemical burns, severe infections, and other corneal or ocular surface disorders. Also among the excluded criteria were patients with incomplete data, poor compliance, pregnant/nursing women, and patients wearing rigid gas-permeable lenses during the last 4wk. Patients were further divided into two subgroups for analysis based on Kmax at postoperative 1mo had improved (Group 1, the initial improvement group)or worsen (Group 2, the initial deterioration group) compared to the preoperative level.

DISCUSSION

In the present study, our overall results revealed that all keratometric parameters achieved reduction of approximately 0.50-1.0 D at 12mo following CXL treatment in keratoconic eyes. It is not surprising that Kmax achieved the most improvement (1.0 D), which is consistent with previous studies

, since Kmax represents the most severe protrusion point and naturally exhibits the most dramatic changes following crosslinking. Ksteep achieved the least reduction(0.47 D). Previous studies reported a reduction of Ksteep (the steepest simulated keratometry) varied from 0.49 to 5.16 D

. It is assumed that the ethnicity difference of the patients, the variety of treatment protocol, the inherent variability in the progression of keratoconus, as well as the limited reproducibility of parameters measurement and the difference of instruments all take responsible for such a large disparity. Besides, it is noteworthy that partial Ksteep in the previous studies may actually represent the Kmax. Ksteep and Kmax represent the two versions of the maximum keratometry, in which the former represents the meridian maximum keratometry within a 3 mm area centered over the corneal apex, while the latter represents the single point maximum keratometry at the anterior surface surrogated for center of the cone. It is reasonable that Kmax captured more prominent changes compared to Ksteep. The apical cornea may exhibit little or mild changes when the conic area protrudes, in decentered cones this phenomenon would be particularly obvious

. Bardan

reported a larger percent of peripheral cones (38.4%) and disclosed that Kmax may not be central although cornea thins centrally. In the present study,there were 82.5% central, 15% paracentral and 2.5% peripheral cones.

Ray tracing method, which takes in account of the anterior and posterior corneal surface and calculates corneal power with the actual refractive index without relying on any assumptions,could overcome the aforementioned limitations and evaluate corneal power more accurately. Besides, Pentacam HR(Oculus Optikger?te GmbH, Wetzlar, Germany), the most widely used corneal tomographer, has provided a customized ray traced corneal power-total corneal refractive power(TCRP), which could calculate certain area centered over certain point. Continued efforts to establish reliable metrics which can accurately assess the effect of CXL is a fundamental requirement for clinical management. To the best of our knowledge, limited information in the literature could prove our conjecture that could TCRP be regarded as a more reliable indicator. Thus, the current study intends to compare TCRP with other commonly used keratometric parameters during 12-month follow-up after CXL and assess the performance of TCRP in evaluating the efficacy of CXL treatment.

At baseline and each time point of the postoperative follow-up examinations (1, 3, 6,and 12mo), all patients underwent uncorrected distance visual acuity (UDVA) and corrected distance visual acuity (CDVA)assessment, non-contact intraocular pressure tonometry, slitlamp biomicroscopy, Pentacam HR corneal tomography(Version 1.21r41). The coordinates of Kmax values (surrogate for center of the cone) were automatically determined by the Pentacam HR software and recorded for each eye at each time point after measurements performed under dim light. In the“Power Distribution” display, a 3 mm zone centered over the corresponding coordinate was manually defined on the sagittal curvature map and the TCRP map at each time point (Figure 1).This procedure was repeated for a 2-mm and 1-mm diameter zone. Two software custom keratometric parameters, including TCRP within the zone centered over Kmax (TCRP

)and mean keratometry within the zone centered over Kmax(Km

) were recorded for analysis, accompanied with keratometric value at the steepest meridian (Ksteep), Kmax,TCRP centered over the apex within a diameter of 3 mm zone (TCRP

3 mm). Only a measurement with quality specification of “OK” was accepted for analysis.

A total of 40 eyes of 30 patients, with 23.3%(7/30) of female were enrolled. UDVA were 0.85±0.36 (range:0.22 to 1.70) and 0.82 (0.40, 0.96) logMAR pre- and 12mo postoperatively (

=0.11); CDVA were 0.22 (0.05, 0.30) and 0.10 (0, 0.30) logMAR (

=0.24); TCT were 478 (467, 500) μm and 473.5 (457.5, 486.5) μm (

＜0.001). Based on the Amsler-Krumeich classification, 28 eyes were defined as stage I, 12 eyes were defined as stage II. Detailed baseline characteristics were in Table 2.

TCRP represents total corneal power by ray tracing method,which takes in account of anterior and posterior corneal surface using actual corneal refractive index. Its measurement principle determines its superiority in evaluating corneal power in altered corneas, such as post-refractive eyes

. Considering local thinning and protrusion in keratoconic eyes disrupted the natural ratio of anterior-to-posterior corneal curvature more seriously, it is reasonable for researchers to begin to shift their interest into TCRP evaluation in eyes with keratoconus.A retrospective study by Takahashhi

disclosed that TCRP

3 mm decreased by 1.0 D at 12mo after CXL similar to the simulated keratometry. In the current study,TCRP

3 mm instead of TCRP

3 mm displayed a smaller reduction of approximately 0.70 D at 12mo after CXL in the overall group analysis. More importantly, TCRP

3 mm and TCRP

1/2/3 mm displayed statistically significant improvement early at 3mo in the overall group,while other keratometric parameters showed statistically significant improvement until 6mo. Besides, TCRP

1 mm/2 mm exhibited an identical improvement compared to Kmax at 12mo. We initially conjectured that TCRP

1 mm/2 mm both could serve as an earlier indicator for evaluating the improvement by CXL treatment during 12mo follow-up. To further verify this speculation, we performed the subgroups analysis and disclosed different results. In Group 1, all the other keratometric parameters disclosed significant improvement as early as 3mo, whereas Kmax, TCRP

2 mm and Km

2 mm detected significant improvement even earlier at 1mo postoperatively. Besides, the improvement exhibited by TCRP

2 mm at each follow-up time point showed no statistically significant difference from Kmax.This further indicates that only TCRP

2 mm could be accepted as an earlier indicator for evaluating the improvement by CXL treatment. Meanwhile, only the keratometric parameters based on ray tracing method and Km

3 mm showed slight improvement as early as 3mo postoperatively in Group 2. TCRP

1 mm achieved the least amount of improvement (0.07 D), which was so faint that could be approximately neglectable in the clinical daily practice. As the defined measurement zone centered over cone became larger,the improvement increased gradually. It is speculated that a relatively more peripheral ectatic region or even normal region of cornea incorporated into the TCRP calculation may exhibit more distinct topographic changes and subsequently influence the TCRP changes following CXL treatment. However, a larger diameter of 4 mm was not included in the present study considering that it may incorporate either more normal cornea or far peripheral or extrapolated data. Although improvement of approximately -0.15 D to -0.20 D obtained with TCRP

2 mm and 3 mm were neither statistically significant nor clinically relevant, this still represents a trend that keratometric parameters based on ray tracing method may detect an earlier improvement compared to the widely used keratometric parameters. Taking all these analyses together, we speculated that TCRP

2 mm could earlier detect the improvement by CXL treatment during 12mo period follow-up. This may represent the most important findings in the current study,that our work firstly explored a newly keratometric parameter which could be used as an early indicator in detecting the effect of CXL on keratoconus stabilization and improvement. The possible mechanisms behind this phenomenon could be that TCPR

takes in account of the neglected posterior corneal surface, captures a certain area instead of a single point, and centers over the cone instead of the apex.

本屆會議以“集成式木工智造·個性化解決方案”為主題，來自中國和世界各地的頂尖行業專家們相聚一起，全面、深入地分享了各自成功的家具和木工機械智造解決方案。

項目來源有多種，如教師的科研項目、某些領域的研究熱點、發明專利，等等。就目前我們開展的項目來看，基本是軟件和硬件結合的項目。項目開發的目標是參加各種競賽及形成科研成果，如發表學術論文、申請專利。……