**2. Materials and Methods**

In this retrospective, cross-sectional study, patients who attended the Department of Ophthalmology, Hiroshima University Hospital, Japan, were evaluated between April 2017 and November 2018. The study received approval from the institution's ethics committee (E—797), and the research adhered to the tenets of the Declaration of Helsinki.

Sixty-one eyes from 50 patients with OHB who had undergone TLE with a fornixbased conjunctival flap using MMC at least 3 months prior to entry were analyzed in this study (OHB group). The eyes with multiple glaucoma surgery were also included in the OHB group. Sixty-five control eyes from 54 glaucoma patients who had no history of prior surgical intervention, except uncomplicated cataract surgery at least 3 months prior to their entry, were concurrently recruited during a similar period (control group). Best-corrected visual acuity (BCVA) and intraocular pressure (IOP; Goldman Applanation Tonometer, Haag-Streit, Köniz, Switzerland) were measured. The anterior segment was observed using slit lamp microscopy. In addition, all eyes with IOLs were using monofocal lenses.

The TOPCON KR-1W wavefront analyzer (Tokyo, Japan) can calculate "corneal" HOAs from the shape of the cornea as well as HOAs of the entire eye. We used the TOPCON KR-1W wavefront analyzer to measure corneal HOAs for a 4 mm pupil diameter without dilating the pupil, and the data were expanded to Zernike polynomials. The magnitude was demonstrated as the mean root square (RMS). Based on our reports, we evaluated corneal wavefront aberrations for coma (C−<sup>1</sup> <sup>3</sup> and C<sup>1</sup> 3), spherical aberrations (C0 4), the RMS of the third-order, fourth-order, and total HOAs. The RMS of the third-order Zernike coefficients (the square root of the sum of the squared coefficients of C−<sup>3</sup> 3, C−<sup>1</sup> 3, C1 3, and C<sup>3</sup> 3) was considered a coma-like aberration. The RMS of the fourth-order Zernike coefficients (the square root of the sum of the squared coefficients of C−<sup>4</sup> 4, C−<sup>2</sup> 4, C<sup>0</sup> 4, C2 4, and C<sup>4</sup> 4) was considered a spherical-like aberration. Finally, the total of HOAs was defined as the RMS of the magnitudes for the third- to fourth-order aberrations. All patients had BCVA of ≥20/40, which was enough to allow fixation on the target of the wavefront analyzer. Aberrometry measurements were automatically measured three times.

For clinical photographic images, we used the TOPCON SL-D8Z slit lamp mounted camera (Tokyo, Japan) to evaluate the anterior ocular segment (×10) from 40 degrees on the temporal side using a diffuser by 50–75 Ws (Watt seconds). OHB was diagnosed if the cornea was covered with bleb under the slit lamp. The dimensional parameters in clinical photographic images were calculated using the NIH image J software (Image J, National Institute of Health, Bethesda, Maryland, USA). The entire corneal area and the area of the bleb over the cornea were measured each as pixels. The ratio of cornea covered by bleb, the ratio of cornea covered by bleb to the entire cornea was calculated as the ratio of the OHB area in the cornea relative to the entire corneal area (Figure 1). We calculated the ratio of the cornea covered by the bleb as follows.

**Figure 1.** Clinical photographic image using slit lamp mounted camera and dimensions measured using NIH image J software. The ratio of cornea covered by bleb is the ratio of area of the cornea covered by bleb to entire cornea area. The red and black lines are the cornea covered by bleb area and entire cornea area, respectively.

The cornea covered by bleb area (pixels)/the entire cornea (pixels) × 100 (%)

The exclusion criteria were BCVA <20/40 and patients with any history of ocular surgery (other than uncomplicated cataract surgery) for the control group. Patients of the OHB group were not excluded for having had glaucoma surgery several times. Additionally, patients of corneal, conjunctival ocular disease observed on a slit-lamp microscopy were also excluded (e.g., pterygium, superficial punctate keratopathy, corneal opacity, and corneal erosion).

#### *Statistical Analysis*

Data were entered in an Excel spreadsheet (Microsoft Corp. Redmond, WA, USA) and analyzed using JMP software (ver. 14, SAS, Inc., Cary, NC, USA). Measurement data were expressed as the mean ± standard deviation with a 95% confidence interval. Continuous data from the two groups were analyzed by an independent *t*-test, whereas discrete data were analyzed using Pearson's chi-square test. The influence of age, IOP, and the ratio of bleb area on aberrations was analyzed using multiple linear regression. BCVA was converted into logMAR units for analysis. Differences were statistically significant when the *p*-value was <0.05.
