Patient Population

This study followed the tenets of the Declaration of Helsinki; the prospective protocol-driven study was approved by the Institutional Review Board of the University of Tokyo School of Medicine, Tokyo, Japan; and written informed consent to participate in research protocol was obtained from all subjects. The observational cross-sectional study population consisted of 26 healthy right eyes of 26 Japanese subjects (20 men and 6 women). The exclusion criteria were presence of ocular diseases, history of intraocular surgeries including refractive surgeries, TBUT of fewer than 5 seconds, Schirmer values of less than 5 mm, instillation of any kind of eye drops in the previous 24 hours, contact lens wear within 2 weeks, and any use of systemic medication.

Tear Meniscus Measurements

Anterior segment imaging using the SS-1000 with custom 16 vertical raster scans at 1-mm intervals was performed for all subjects. This custom scan mode, which we referred to as the tear meniscus mode, was used to obtain a total of 16 cross-sectional images of the anterior segment with width of 16 mm and length of 16 mm. A scan with angle of 90 degrees visualizes both upper and lower tear meniscus simultaneously ( Figure 1 ). The acquisition time for all scans is 0.3 seconds. The OCT image consists of 512 A-scans, with an axial resolution of approximately 8 μm and a transverse resolution of approximately 30 μm. Subjects were asked to look straight at a fixation light with no background illumination and were allowed to blink spontaneously during the examination. The right eyes were examined, but the contralateral eyes were maintained open while capturing the image. All subjects underwent OCT examination 1 second after blinking. Stable ambient room light was maintained during the examination. After 15 minutes, the same examiner (R.F.) repeated the scan for 17 subjects who consented to repeat the examination procedure. The upper and lower tear meniscus height, upper and lower tear meniscus area, and lower tear meniscus volume were evaluated using the SS-1000 by 2 masked graders (R.F.,T.U.).

Tear meniscus image obtained by SS-1000 optical coherence tomography (Tomey Corp, Nagoya, Japan) in a normal subject. A vertical 16-mm scan across the corneal apex was obtained 1 second after a normal blink. The upper eyelid, lower eyelid, cornea, upper tear meniscus, and lower tear meniscus are visualized in a single image.

First, the original OCT images were magnified by 300% to measure the tear meniscus. Both the upper and lower tear meniscus height values were measured manually using the software calipers on the vertical scan image across the central cornea. The central cornea in this case refers to the apex of the cornea, in which this OCT focuses automatically while capturing the image. Second, outlines of both the upper and lower tear meniscus at the central cornea, which consisted of the corneal surface, lid margin, and tear meniscus surface, were plotted manually using the software calipers ( Figure 2 ). At least 9 points were used for the plotting of each image. The area within the plotted lines, the tear meniscus area, was calculated automatically by the program, and the resulting tear meniscus area value was divided by the refractive index of balanced salt solution (1.343) to correct the refraction at the air–meniscus interface. Finally, the central 11 images of 10 mm in width were selected among the 16 images. The lower tear meniscus area of each image was identified as mentioned above, and the tear meniscus area of 11 consecutive images was integrated automatically by the program to calculate the tear meniscus volume. Again, the calculated value was corrected using 1.343 ( Supplemental Figure at AJO.com ). Because the program required consecutive images to calculate the volume, the upper tear meniscus volume could not be evaluated because the presence of eyelashes rendered it difficult to visualize the entire upper tear meniscus adequately.

Image showing 300% magnification of a lower tear meniscus image obtained by swept-source optical coherence tomography. The horizontal line indicating the tear meniscus height. The area within the manually plotted lines indicates the tear meniscus area. The program calculates the tear meniscus volume as the integral of tear meniscus areas identified in 11 consecutive images.

Examinations

The other ophthalmic examinations were slit-lamp microscopy, fluorescein staining, TBUT, and Schirmer test without local anesthesia. A fluorescein sodium strip was wetted with a drop of nonpreservative buffered saline solution and was instilled in the inferior palpebral conjunctiva. The TBUT measurement was obtained twice, and the mean value of the measurements was calculated. Fluorescein staining of the ocular surface also was recorded as described previously. Fluorescein staining scores ranged between 0 and 9 points. Five minutes after the TBUT test, the Schirmer test was performed without anesthesia. A dry Schirmer test strip (Showa Yakuhin Kako Co, Tokyo, Japan) was placed over the lower lid margin on the conjunctival sac at the lateral one third of the eyelid. After 5 minutes with the subject’s eyes closed, the strip was removed and the amount of wetting was recorded in millimeters.

Statistical Analysis

The intergrader repeatability and interimage repeatability of the tear meniscus height, area, and volume were analyzed. Further, the intraclass correlation coefficient (ICC) and coefficient of variation (CV) for the tear meniscus measurements were evaluated. ICC has been advocated as a statistic for assessing the agreement or consistency between 2 methods of measurement. The CV shows the extent of variability in relation to the mean of the population. It is defined as the ratio of the standard deviation to the mean, multiplied by 100 to be expressed as a percent. Correlations between the tear meniscus measurements and tear fluid parameters including TBUT and Schirmer values were analyzed using the Spearman nonparametric correlation analysis.