Vitrectomy combined with various manipulations of the internal limiting membrane (ILM) such as ILM peeling, inverted flap, and autologous flap results in successful closure of the macular hole (MH) in almost all cases.1–4 However, functional recovery is less satisfactory compared to the anatomical success rate.5,6 Postoperative metamorphopsia particularly remains an unresolved complication.7
Recently, it was reported that the removal of the ILM induced macular deformation.8 Importantly, the extent of the removal was associated with the degree of metamorphopsia.9 However, debates remain regarding the mechanisms of the macular deformation and metamorphopsia.8,9 This may be due to that most studies investigating postoperative deformation were based on the inner retina, since the assessment of the outer retina is less feasible due to the lack of landmarks. Those results could not provide information regarding the photoreceptor layer and fovea, which are more directly related to visual function.
The progress of enhanced depth imaging and swept-source optical coherence tomography (SS-OCT) enabled visualization of the choroidal vessels in en face images. The choroidal vessels provide a reference for postoperative displacement of the fovea, as MH surgery would not deform the choroid.
Herein, we investigated the postoperative displacement of the foveal retinal layers using en face OCT images after MH surgical closure.
Patients and Methods
This study was a retrospective, interventional case series. We included idiopathic patients with MH who underwent vitrectomy combined with the removal of ILM in Pusan National University Hospital and Pusan National University Yangsan Hospital from 2014 to 2016. All patients were diagnosed with a full-thickness MH and were followed up for at least 6 months postoperatively. Informed consent was obtained from all patients regarding the risks and benefits of the surgery. Patients were excluded if they exhibited secondary MH, retinal detachment, diabetic retinopathy, macular degeneration, epiretinal membrane, media opacity, high myopia (spherical equivalent ≥ −6.0 diopters or axial length ≥ 26 mm), operation using autologous ILM flap, or other ocular conditions that cause bias in the interpretation of foveal displacement.
The institutional review board of Pusan National University Hospital approved the study design (approval No. 1706-005-056). The present study adhered to the tenets of the Declaration of Helsinki.
A three-port pars plana vitrectomy was performed using the Constellation (Alcon Laboratories, Fort Worth, TX) 25-gauge vitrectomy system by three surgeons (SWB, ISB, and JEL). Phacoemulsification with intraocular lens implantation was combined concurrently at the surgeon's discretion. After the core and posterior vitreous were removed, the ILM was removed from an area of approximately 3 disc diameters (DD) at the center of the macula with assistance of triamcinolone acetonide or vital stain. Room air or 18% sulfur hexafluoride was used to fill the vitreous cavity, and all patients were instructed to maintain a face-down position for 1 to 3 days.
All patients underwent comprehensive ophthalmologic examinations at baseline and at 1, 3, and 6 months' follow-up, including best-corrected visual acuity (BCVA), fundus examination, fundus photography (Canon CR-2 digital non-mydriatic retinal camera; Canon, Tokyo, Japan), and SS-OCT (DRI OCT-1 Atlantis; Topcon, Tokyo, Japan).
Foveal Displacement Measurements
To assess the foveal displacement, distinct hyperreflective bands were reviewed. The foveal pit was inconsistently seen at the outer plexiform layer and retinal nerve fiber layer. In the surgically closed MH, foveal depression was as deep as to the outer nuclear layer in some eyes, but not in the others. Meanwhile, due to asymmetric thickening of the nasal and temporal macula, the shape of the foveal pit was also inconsistent in the en face image at the level of the retinal nerve fiber layer. Accordingly, the ellipsoid zone (EZ) and inner plexiform layer (IPL) were chosen for the analysis.
Figure 1 demonstrates the following measurement schematically. OCT scans were obtained by using the three-dimensional (3-D) macular scan protocol (6 mm × 6 mm, 512 × 256 scan) at baseline and at 1, 3, and 6 months postoperatively. En face OCT images were exported at the level of the choroid, EZ, and IPL from the same volume scan at each visit. The reference plane of Bruch's membrane (BM) segmentation was manually moved to each layer. The layer of the choroid was selected to show a large blood vessel, which can serve as a marker in the en face image. The EZ and IPL layers were determined based on hyperreflective layers in the cross-sectional image.
This method was used for measuring the displacement of the foveal center. The center of the ellipsoid zone (EZ) was marked in the en face image by reviewing all B-scans from optical coherence tomography (OCT) volume scanning. The center of the inner plexiform layer (IPL) was designated as the foveal depression shown in the en face image at the level of the IPL. Both centers were marked on the choroid en face image obtained from the same volume scan. These processes were performed using the OCT volume scan obtained at every visit. The en face images of the choroid were stacked to fit the choroidal vasculature pattern. Coordinates of the foveal centers were measured.
The measurement of the foveal displacement was performed using a raster graphics editor (Photoshop; Adobe Systems, San Jose, CA) (Figure 1). The set of exported en face images (choroid, EZ, and IPL) was stacked as layers. The measurement of X coordinates was reversed in the left eye to ensure consistent directionality, and an increase in the X coordinate indicated nasal displacement.
At baseline, the presumed center of the EZ and IPL was defined as the crossing point of the largest dimension of the hole, shown as a round hyporeflective area in the en face image at each layer. After hole closure, the EZ center was defined as a small irregularity or defect of the EZ, indicating the center of the centripetal closure, by reviewing the series of the B-scans (Figure 1). The center of IPL was defined as the center of the foveal pit at the IPL layer in the en face image. The foveal center of the EZ and IPL was marked in the choroidal en face image obtained from the same volume scan.
The marked choroidal images were overlapped to match the vasculature, and the center at baseline was set as (0,0) of X and Y coordinates. The postoperative location of the foveal center was measured at each visit as X and Y coordinates, and the amount and direction were calculated as vector.
The en face image was used to determine the size of ILM that was removed. By setting the reference plane as the ILM segmentation, a demarcated circular defect could be identified, representing the area of ILM removal. This area was measured horizontally and vertically by using software mounted in the SS-OCT (IMAGEnet 6; Topcon, Tokyo, Japan). If the area of ILM removal was indiscernible in the en face image, a red-free image was used for the measurement.
The aperture diameter of MH was defined as the minimum dimeter of the hole, and the base diameter of MH was defined as the diameter of the hole at the EZ layer.10
Measurement of the Vascular Displacement
The displacement of the retinal vessels was evaluated by comparing the fundus photographs using a customized software as reported previously.8 Sets of fundus photographs were taken at baseline and at 6-month follow-up. A set of two photographs was overlapped. The angle, scale, and position of the postoperative photograph were adjusted to the baseline photograph using four alignment vectors calculated from the difference of location at the vascular bifurcations. An en face OCT image was utilized to calculate the scale of the photograph by matching two points of vascular bifurcation between the en face OCT and preoperative photograph. The vascular displacement was measured in 16 sectors composed of two rings that were divided into eight sectors; temporal, superotemporal, superior, superonasal, nasal, inferonasal, inferior, and inferotemporal. The diameters of the inner and outer rings were 2 mm to 4 mm and 4 mm to 6 mm, respectively, and a central area of 2 mm was excluded due to vascular paucity.
BCVA was measured using the Snellen chart and was converted to the logarithm of minimum angle of resolution (logMAR) scale for statistical analysis. Normality tests were performed using the Kolmogorov-Smirnov and Shapiro-Wilk analyses. Changes in the BCVA, X and Y coordinates of the fovea, and difference of the foveal center location between the retinal layers and vascular displacements were analyzed using the Wilcoxon signed-rank test. Correlations of the foveal location in the EZ and IPL, and between foveal displacement and other parameters such as vascular displacement or various baseline characteristics were estimated using Spearman's correlation analysis. Categorical variables including dye for ILM staining, tamponade material, lens status, and sex were evaluated by a chi-square test or Fisher's exact test. Associations of the removal of ILM with displacements were evaluated using univariate linear regression analyses. The statistical analyses were performed using the SPSS version 22.0 for Windows software (IBM, Armonk, NY). P values less than .05 were considered statistically significant.
A total of 26 eyes (seven men and 19 women) were included in the study. Table 1 summarizes the patients' baseline characteristics. The mean age was 64.2 years ± 6.9 years, and mean BCVA was 0.65 logMAR (20/90) ± 0.25 logMAR at baseline. Lens status was phakic in 22 eyes and pseudophakic in four eyes. Combined phacoemulsification with intraocular lens implantation was performed in 21 eyes and the lens was spared in one eye, which did not have any significant progression of cataract to affect the image quality during the observation period. The mean aperture diameter was 333.1 μm ± 150.5 μm and mean basal diameter was 594.3 μm ± 255.0 μm. As an adjuvant for the ILM staining, triamcinolone acetonide was used in four eyes, indocyanine green in 13 eyes, and brilliant blue G in nine eyes. Size of ILM removal was measured in 24 eyes (20 in en face images and four in red-free images), and the mean size was 4,708.0 μm ± 637.0 μm (range: 3,723 μm – 6,094 μm).
Baseline Characteristics of Idiopathic Macular Hole
MH was closed in all 26 eyes. BCVA improved significantly to 0.49 logMAR (20/63) ± 0.26 logMAR at 1 month (P = .003), 0.34 logMAR (20/43) ± 0.15 logMAR at 3 months (P < .001), and 0.23 logMAR (20/34) ± 0.15 logMAR at 6 months (P < .001).
Foveal Displacement in EZ and IPL
Table 2 summarizes the mean foveal displacement distance and angle in the EZ and IPL. Mean displacements at 1, 3, and 6 months' follow-up were 53.9 to −12.0°, 116.1 to −14.3°, and 118.2 to −12.6° in the EZ and 75.5 to −5.4°, 117.9 μm to −6.6° and 137.4 to −6.7° in the IPL, respectively. The foveal centers of the EZ and IPL were displaced to the disc significantly in the X coordinate during all follow-up periods (P < .001). The displacement occurred mostly in the first 3 months and exhibited statistical significance at 3 months compared with 1 month in the EZ (P = .001) and IPL (P = .009), but not at 6 months compared with 3 months. Displacements of the fovea in the Y coordinate were significantly inferior in the EZ at 3 and 6 months (P = .006 and P = .017, respectively), but not in the IPL. At 3 and 6 months, displacement was significant compared with the previous follow-up in EZ (P = .022 and P = .035, respectively).
Fovea Center Displacement
The IPL center was located slightly superior to the EZ center during follow-up (Table 2, Figure 2). No difference was observed in the foveal center location between the EZ and IPL (P = .086, P = .552, and P = .188 on the X coordinate and P = .829, P = .210, and P = .144 on the Y coordinate at the 1, 3, and 6 months, respectively). The X and Y coordinates of the EZ center showed a significant correlation with the coordinates of the IPL center at 1, 3, and 6 months (P ≤ .016; Spearman's correlation coefficient = 0.496–0.860).
The diagram shows the relative location of the centers of the ellipsoid zone (EZ) and inner plexiform layer (IPL) after the surgical closure of the macular hole, compared with the center of the EZ at baseline. The X and Y coordinates of two centers were correlated significantly (P ≤ .016, Spearman's ranked correlation).
The degree of EZ and IPL foveal displacements did not show a correlation with any baseline parameters including the BCVA, MH diameter, adjuvant for ILM peeling, removal size of ILM, tamponade material, sex, age, spherical equivalent, axial length, and lens status.
Vascular Displacement in the Fundus Photo Analysis
The mean vascular displacement distance was 53.9 μm to −5.1° in the whole sector, 80.1 to −3.4° in the inner ring, and 27.9 to −10.1° in the outer ring at 6 months. The mean displacement distance and angle in each sector are summarized in Table 3. The displacement was the largest (149.8 μm) in the temporal sector of the inner ring and the smallest (19.1 in the inferior sector of the outer ring. The displacement was significantly larger in the inner ring than the outer ring, and in the temporal sectors than the nasal sectors.
Distance and Angle of the Mean Retinal Vessel Displacement in Each SectorAfter MH Closure
The horizontal displacement of the EZ center was correlated with the vascular displacement in the temporal sector of the inner ring (P = .014), and the vertical displacement of the EZ center was correlated with the vascular displacement in the superior and inferior sectors of the inner ring (P = .029 and P = .024).
The present study investigated the displacement of the foveal photoreceptor layer using en face OCT images after vitrectomy combined with ILM peeling in idiopathic MH and highlighted concurrent displacement of the outer and inner retina to the optic disc with a significant correlation. Foveal displacement occurred mostly within 3 months and had a correlation with the vascular displacement at 6 months. Asymmetric vascular displacement was demonstrated in agreement with previous studies.7–9
As there are no retinal layers in MHs before operation, we presumed that center of the EZ and IPL was the center of the hole depicted in the en face image. The correspondence between the baseline location and the inverse estimation of postoperative displacements shown in Figure 2 support our presumption.
Regarding the postoperative displacement of the macula after removal of the ILM, two different mechanisms were proposed. Bae et al.9 claimed that contraction of the residual ILM accounted for the asymmetric elongation of the parafoveal tissue, whereas other investigators indicated that contraction of the retinal nerve fiber layer (RNFL) causes displacement of the sensory retina untethered from the rigid ILM.11–13 For the measurement of vascular displacement, we compared pre- and postoperative fundus photographs. When the en face image of the ILM segmentation was overlapped, it clearly demonstrated that vascular displacement occurred mostly where the ILM had been removed (Figure 3). If the epiretinal membrane developed on the residual ILM, it may cause contraction of the underlying retina. Otherwise, contraction of the ILM did not appear to cause significant contraction.
This comparison of pre- and postoperative fundus photographs demonstrates vascular displacement within the border of the internal limiting membrane (ILM) removal. The fundus photographs were obtained at (A) baseline and (B) postoperative 6 months. (C) En face image of ILM segmentation at 6 months shows the margin of the ILM peeling. (D) Pre- and postoperative fundus photographs are overlaid, and the area of the ILM removal margin is marked on the composite photograph. Displacement is distinct in the vessels where the ILM was removed.
The fovea lacks the inner retina and does not have the RNFL. Consequently, the force of the RNFL contraction would displace the fovea indirectly by displacing the parafoveal tissue. These are supported by the distance of foveal displacement. The foveal displacement was less than the inner temporal sector and greater than the inner nasal sector. In addition, the vascular displacement of the inner ring was correlated with the displacement of the EZ center. Similarly, we speculated that the inner retinal contraction caused indirect outer retinal displacement. The center of the IPL was slightly superonasal from the center of the EZ after surgical closure, and they were displaced concurrently with a significant correlation (Table 2, Figure 2). The superonasal direction is toward the optic disc and would be the sum of the force vector generated by the RNFL contraction.
In previous reports, the MH diameter and extent of ILM removal were related to the distance of the vascular displacement.8 However, no parameters were found to be associated with foveal displacement in the present study. The association of the MH diameter with the vascular displacement was related to centripetal movements of the parafoveal tissue in the process of the hole closure. Accordingly, the MH diameter would not be related to the location of the foveal center, which is the center of the centripetal contraction. Regarding the extent of the ILM removal, the present study included more recent cases where the ILM was removed mostly in about 3-DD area. The distribution of the area that was removed appeared too narrow to show a correlation between foveal displacement and area size.
The mechanism of metamorphopsia is still unclear, especially in macular diseases including epiretinal membrane and MH without subretinal pathology. The simplest theory is that changes in the arrangement of the photoreceptors may deform the visual perception.14,15 However, there is no study to date that evaluated the spatial relationship between the inner and outer retina. It was also hypothesized that the status of the inner retina would account for metamorphopsia.6 The role of the Müller cells as optical fibers and disturbance in the synaptic junctions were proposed as possible mechanisms.16–18 Our results showing the concurrent displacement of the photoreceptor and inner plexiform layer supported the former hypothesis.
Asymmetric elongation of the fovea was associated with postoperative metamorphopsia.6,19 Interestingly, a narrower removal of the ILM was related to more asymmetric elongation and more metamorphopsia.9 Based on the results of the present study, the associations can be explained as follows. When the ILM is removed in a narrow area, the bending point of the displacement is close to the foveal center and the parafoveal tissue is arranged asymmetrically, thus accounting for metamorphopsia. In contrast, in a wide ILM removal, foveal and parafoveal tissues are displaced altogether and the bending point is farther from the foveal center, causing less metamorphopsia. To verify this theory, further investigation is warranted.
There are several limitations in this study. This was a retrospective study with a short observation period, and the sample size was too small to find factors related to the foveal displacement. Metamorphopsia examination was not performed, and the relationship between postoperative displacement and metamorphopsia was not evaluated. Therefore, future studies should be conducted to determine the relationship between postoperative metamorphopsia and deformations of the parafoveal tissue.
In summary, the results of our study confirmed, using a fixed reference under the fovea for locating the foveal retinal layers, that the photoreceptor layer was displaced to the disc concurrently with the inner retina after vitrectomy with ILM peeling in idiopathic MH. It is mandatory to perform the future study to investigate the correlation between postoperative metamorphopsia and deformed arrangement of the photoreceptors to determine mechanism of metamorphopsia.
- Brooks HL Jr, . Macular hole surgery with and without internal limiting membrane peeling. Ophthalmology. 2000;107(10):1939–1948. doi:10.1016/S0161-6420(00)00331-6 [CrossRef]
- Michalewska Z, Michalewski J, Adelman RA, Nawrocki J. Inverted internal limiting membrane flap technique for large macular holes. Ophthalmology. 2010;117(10):2018–2025. doi:10.1016/j.ophtha.2010.02.011 [CrossRef]
- Shin MK, Park KH, Park SW, Byon IS, Lee JE. Perfluoro-n-Octane–assisted single-layered inverted internal limiting membrane flap technique for macular hole surgery. Retina. 2014;34(9):1905–1910. doi:10.1097/IAE.0000000000000339 [CrossRef]
- Morizane Y, Shiraga F, Kimura S, et al. Autologous transplantation of the internal limiting membrane for refractory macular holes. Am J Ophthalmol. 2014;157(4):861–869. doi:10.1016/j.ajo.2013.12.028 [CrossRef]
- Itoh Y, Inoue M, Rii T, Hiraoka T, Hirakata A. Correlation between length of foveal cone outer segment tips line defect and visual acuity after macular hole closure. Ophthalmology. 2012;119(7):1438–1446. doi:10.1016/j.ophtha.2012.01.023 [CrossRef]
- Arimura E, Matsumoto C, Okuyama S, Takada S, Hashimoto S, Shimomura Y. Quantification of metamorphopsia in a macular hole patient using M-CHARTS. Acta Ophthalmol Scand. 2007;85(1):55–59. doi:10.1111/j.1600-0420.2006.00729.x [CrossRef]
- Kim JH, Kang SW, Park DY, Kim SJ, Ha HS. Asymmetric elongation of foveal tissue after macular hole surgery and its impact on metamorphopsia. Ophthalmology. 2012;119(10):2133–2140. doi:10.1016/j.ophtha.2012.05.018 [CrossRef]
- Pak KY, Park KH, Kim KH, et al. Topographic changes of the macula after closure of idiopathic macular hole. Retina. 2017;37(4):667–672. doi:10.1097/IAE.0000000000001251 [CrossRef]
- Bae K, Kang SW, Kim JH, Kim SJ, Kim JM, Yoon JM. Extent of internal limiting membrane peeling and its impact on macular hole surgery outcomes: A randomized trial. Am J Ophthalmol. 2016;169:179–188. doi:10.1016/j.ajo.2016.06.041 [CrossRef]
- Duker JS, Kaiser PK, Binder S, et al. The International Vitreomacular Traction Study Group Classification of Vitreomacular Adhesion, Traction, and Macular Hole. Ophthalmology. 2013:120(12):2611–2619. doi:10.1016/j.ophtha.2013.07.042 [CrossRef]
- Ishida M, Ichikawa Y, Higashida R, Tsutsumi Y, Ishikawa A, Imamura Y. Retinal displacement toward optic disc after internal limiting membrane peeling for idiopathic macular hole. Am J Ophthalmol. 2014;157(5):971–977. doi:10.1016/j.ajo.2014.01.026 [CrossRef]
- Nakagomi T, Goto T, Tateno Y, Oshiro T, Iijima H. Macular slippage after macular hole surgery with internal limiting membrane peeling. Curr Eye Res. 2013;38(12):1255–1260. doi:10.3109/02713683.2013.811261 [CrossRef]
- Yoshikawa M, Murakami T, Nishijima K, et al. Macular migration toward the optic disc after inner limiting membrane peeling for diabetic macular edema. Invest Ophthalmol Vis Sci. 2013;54(1):629–635. doi:10.1167/iovs.12-10907 [CrossRef]
- Krøyer K, Christensen U, la Cour M, Larsen M. Metamorphopsia assessment before and after vitrectomy for macular hole. Invest Ophthalmol Vis Sci. 2009;50(12):5511–5515. doi:10.1167/iovs.09-3530 [CrossRef]
- Saito Y, Hirata Y, Hayashi A, Fujikado T, Ohji M, Tano Y. The visual performance and metamorphopsia of patients with macular holes. Arch Ophthalmol. 2000;118(1):41–46. doi:10.1001/archopht.118.1.41 [CrossRef]
- Franze K, Grosche J, Skatchkov SN, et al. Müller cells are living optical fibers in the vertebrate retina. Proc Natl Acad Sci U S A. 2007;104(2):8287–8292. doi:10.1073/pnas.0611180104 [CrossRef]
- Hibi N, Ueno S, Ito Y, Piao CH, Kondo M, Terasaki H. Relationship between retinal layer thickness and focal macular electroretinogram components after epiretinal membrane surgery. Invest Ophthalmol Vis Sci. 2013;54(12):7207–7214. doi:10.1167/iovs.13-12884 [CrossRef]
- Ichikawa Y, Imamura Y, Ishida M. Metamorphopsia and tangential retinal displacement after epiretinal membrane surgery. Retina. 2017;37(4):673–679. doi:10.1097/IAE.0000000000001232 [CrossRef]
Baseline Characteristics of Idiopathic Macular Hole
|Age (Years, Mean ± SD; Range)||64.2 ± 6.9; 52 – 85|
|Lens Status (Phakic/Pseudophakic)||22/4|
|Combined Phacoemulsification With IOL (No. of Eyes)||21|
|Aperture Diameter (μm, Mean ± SD; Range)||333.1 ± 150.5; 142.5 – 705|
| Vertical diameter||322.4 ± 164.5; 107 – 670|
| Horizontal diameter||343.7 ± 145.2; 169 – 740|
|Base Diameter of MH (μm, Mean ± SD; Range)||594.3 ± 255.0; 160.5 – 1129|
| Vertical diameter||583.7 ± 252.4; 95 – 1116|
| Horizontal diameter||604.8 ± 279.7; 66 – 1142|
|Size of the ILM Removal (Mean ± SD; Range)||4,708.0 ± 637.0; 3,723.5 – 6,094.8|
| Vertical diameter||4,624.2 ± 636.2; 3,469 – 6,033.4|
| Horizontal diameter||4,791.7 ± 785.0; 3,784 – 6308.4|
|BCVA (logMAR, Mean ± SD; Range)||0.65 ± 0.25; 0.3 – 1.4|
|Mean AL (mm, Mean ± SD; Range)||23.5 ± 0.9; 21.4 – 25.2|
|Mean Spherical Equivalent (Diopters, Mean ± SD; Range)||−0.37 ± 1.23; −3.25 – 1.75|
Fovea Center Displacement
|Baseline||1 Months||3 Months||6 Months|
|EZ Center From Baseline||X coordinate (μm)||0||52.7ab ± 42.9||112.5ab ± 69.4||115.4a ± 74.8|
|Y coordinate (μm)||0||−11.2 ± 39.8||−28.7 ab ± 64.0||−25.9 ab ± 72.4|
|IPL Center From Baseline||X coordinate (μm)||0||75.2 ab ± 82.3||117.1 ab ± 111.6||136.5a ± 115.0|
|Y coordinate (μm)||0||−7.1 ± 56.5||−13.5 ± 56.6||−16.0 ± 71.4|
|IPL Center Compared to EZ Center at Baseline||X coordinate (μm)||−1.4 ± 94.9||73.8 ± 66.2||116.3 ± 84.3||135.1 ± 96.4|
|Y coordinate (μm)||22.8 ± 68.4||15.6 ± 72.9||11.4 ± 81.9||6.8 ± 86.4|
|Difference of IPL Center From EZ Center at Each Visit||X coordinate (μm)||−1.4 ± 94.9||21.1 ± 58.3||3.8 ± 76.5||19.7 ± 77.5|
|Y coordinate (μm)||22.8 ± 68.4||26.9 ± 56.5||40.0 ± 48.5||32.7 ± 44.1|
|Foveal Depression Area at IPL Level||Area (mm2)||0.230||0.206||0.240||0.235|
Distance and Angle of the Mean Retinal Vessel Displacement in Each SectorAfter MH Closure
|Inner Ring||Distance (µm)||149.8||99.3||77.4||58.8||43.9||72.1||85.9||105.1|
|Outer Ring||Distance (µm)||49.4||26.0||19.1||20.1||24.0||31.3||33.0||30.0|