Patients and Methods
Nineteen patients with significantly dry eyes were recruited in a consecutive manner from a tertiary cornea practice for the current prospective study. The lower tear meniscus of the right eye in each subject was imaged by vertical scans centered on the inferior cornea and the lower eyelid using an FD-OCT system (RT-Vue software version 4.7; Optovue, Inc., Fremont, CA) with a corneal adaptor. Patients were asked to refrain from placing any lubricating drops or medications for 2 hours prior to their measurements. All of the measurements were performed by one technician. This study was in accordance with the Health Insurance Portability and Accountability Act of 1996. The lower meniscus height and depth were measured with a computer caliper. The cross-sectional area was calculated using a two-triangle approximation. All of the computer caliper measurements were made by the first author.
An FD-OCT system (RTVue) with a corneal adaptor module was used. The system operated at an 830-nm wavelength and had an axial resolution in the tissue of 5 μm. The corneal adaptor module produced telecentric scanning for anterior segment imaging using either a wide-angle or high-magnification adaptor lens. We used the wide-angle lens, which provided a transverse resolution of 15 μm. The room temperature was set at 70°F and humidity was at 40%. Patients were asked to look straight ahead at the fixating target within the OCT system. The OCT pattern used to scan the lower tear meniscus was a 6-mm vertical line centered on the middle of the inferior corneal limbus (Fig. 1). Subjects were instructed to blink and then keep the eyes open for the duration of a 3-second count. Images were taken at 2 seconds after blink.
Figure 1. Video image illustrating the 6-mm vertical scan path (arrow) of the optical coherence tomography beam used to image the lower tear meniscus.
Two baseline measurements were taken for each subject prior to administration of a drop of artificial tears (carboxymethylcellulose sodium 0.5%, Optive; Allergan, Inc., Irvine, CA). Five serial pairs of measurements were then taken at 1, 2, 5, 10, and 15 minutes after the instillation of artificial tears. Only right-eye images were used for the current study. The OCT images were exported for computer caliper measurements of lower tear meniscus height, depth, and cross-sectional area (Fig. 2). The height was measured from the cornea–meniscus junction to the lower eyelid–meniscus junction. The depth was measured from the midpoint of the air–meniscus interface to the cornea–lower eyelid junction. The area was approximated by two triangles (Fig. 2). The saline group index of 1.342 at 830-nm wavelength was used to correct the refraction at the air–meniscus interface.
Figure 2. The tear meniscus height and depth on a vertical optical coherence tomography section centered on the inferior corneal–eyelid junction. The cross-sectional area was calculated using a dual triangle approximation.
The Wilcoxon signed rank test was used to compare tear meniscus height, depth, and area at baseline to each subsequent interval after instillation of artificial tears (1, 2, 5, 10, and 15 minutes). The Wilcoxon signed rank test was specifically used because of our small population size of 16 subjects and the non-normal distribution of our measurements.
The study included the right eye of 19 subjects with dry eye from a tertiary corneal practice. Three subjects were excluded from the data analysis because they had significant conjunctivochalasis precluding accurate measurement of the lower tear meniscus (Fig. 3). The remaining 16 subjects with dry eye reflected the normal predominance of dry eye among the female population, with 10 women and 6 men (age: 52.3 ± 16.1 years).
Figure 3. Lax conjunctival folds prevent accurate measurement of tear meniscus.
The baseline meniscus measurements were 235.5 ± 150.0 μm, 138.1 ± 78.7 μm, and 0.020 ± 0.022 mm2 for height, depth, and area, respectively (Table). At 1 minute after instillation of artificial tears, there was a more than three-fold increase in all lower tear meniscus parameters. Specifically, the increases were 374%, 346%, and 950% for height, depth, and cross-sectional area, respectively. At 5 minutes, there continued to be a statistically significant elevation of all parameters including height (163%), depth (163%), and area (225%) (Table; Fig. 4). By 10 minutes, all of the parameters, had returned to just slightly higher values than baseline. Although there was a trend, there was no statistical difference between baseline and 10 minute values (P > .05).
Table 1: Serial Tear Meniscus Parameters After Artificial Tear Instillation
Figure 4. Lower tear meniscus dynamics after artificial tear instillation.
Dry eye evaluation has traditionally been performed by a series of invasive tests such as Schirmer test, cotton thread test, rose bengal staining, fluorescein staining, and tear break-up time. These tests are considered invasive because they involve contact with the ocular surface or the instillation of a dye. By introducing an irritant, they induce reflex tearing and thus may not accurately reflect basal tear secretion in the dry eye state.1–5
Some authors have switched to using lower tear meniscus measurements to quantify dry eye. Lower tear meniscus parameters have been shown to correlate well with dry eye. Mainstone et al.2 measured height and radius of tear meniscus with a slit-lamp–equipped micrometer. Their findings demonstrated a significant difference between normal subjects and patients with dry eye with lower tear meniscus height having a greater than 90% sensitivity for identifying tear-deficient dry eye.2 Similarly, Kawai et al.7 used a slit-lamp photograph to measure tear meniscus height after instillation of fluorescein and found significantly lower tear meniscus in subjects with dry eye as compared with the normal population. However, the findings of both of these studies may be confounded by the addition of fluorescein required for photographic evaluation.
Reflective meniscometry is a tear meniscus assessment tool that has been developed as an attempt to quantify meniscus volume in a non-invasive manner.8–10 It has been used to study tear dynamics in both normal patients and the population with dry eye.8–10 Although it has been useful in documenting changes in tear meniscus curvature, it does not directly measure tear meniscus height or area. Instead, it requires mathematical models and several assumptions to estimate meniscus volume and cross-sectional area.
With the refinement of OCT, researchers have been able to directly quantify tear meniscus parameters. Shen et al.11 used real-time OCT to measure upper and lower tear menisci in patients with dry eye. They found that lower tear meniscus height was a good predictor of dry eye and had improved sensitivity and specificity as compared to the superior meniscus.11 Quantification of the upper meniscus has been limited by several factors, including eyelashes that block the view of the tear meniscus, rapid movement and twitches of the upper eyelid, and smaller meniscus volumes as compared with the lower meniscus.11,12
In review of the literature, baseline lower tear meniscus measurements have varied depending on both the method used to evaluate the meniscus and the population studied. Slit-lamp video photography has been used to measure tear meniscus height as low as 171 μm in the elderly population.13 Oguz et al.14 used a slit-lamp micrometer to measure tear meniscus height of 190 μm in a population with dry eye. In contrast, Mainstone et al.2 measured elevated tear meniscus heights of 461 μm with a slit-lamp micrometer in the normal population. There has been significant variability between different populations in different studies. However, Johnson and Murphy12 found good agreement between OCT and video measurements of tear meniscus height in their normal population. Lower tear meniscus cross-sectional area has not been given much attention in the past because, prior to the development OCT, there was not a reliable way of directly measuring cross-sectional area.
When compared with normative OCT tear meniscus studies, our baseline dry eye tear meniscus measurements of 236 μm height and 0.020 mm2 area fell within accepted ranges.6,11,12,15–22 Previous OCT studies of normal patients have reported mean tear meniscus heights ranging from 190 to 310 μm, whereas cross-sectional area has ranged from .0150 to .0307 mm.2,6,11,12,15–22 However, our baseline meniscus measurements were higher than previous OCT studies in the population with dry eye. Shen et al.’s11 OCT study on patients with aqueous tear deficiency reported a lower tear meniscus height of 196 μm and a much lower cross-sectional area of .0095 mm2. Similarly, Wang et al.22 reported dry eye values of 179 μm and .01 mm2 for tear meniscus height and area, respectively, in their FD-OCT study. Although slit-lamp photography of the tear meniscus cannot comment on cross-sectional area, it was the primary tool used to measure tear meniscus prior to the development of OCT. When considering the population with dry eye, both our current study results, as well as those of Shen et al. and Wang et al., compare well with the previous slit-lamp studies that reported lower tear meniscus heights ranging from 170 to 290 μm.1,2,7,14 The differences between our current study and Shen et al.’s and Wang et al.’s studies can likely be explained by our different patient populations. Both of these studies11,22 defined their dry eye population by a Schirmer test of less than 5 mm, and thus limited their population to dry eye secondary to severe aqueous tear deficiency. Our dry eye population was a more heterogenous population that included some patients with reflex tearing and Schirmer scores of greater than 20 mm. The reflex tearing in our dry eye population likely accounted for higher baseline meniscus measurements.
In the current study, we aimed not only to examine baseline tear meniscus values, but also to describe tear film dynamics. After instillation of artificial tears, there was a more than threefold initial increase in all lower meniscus parameters that remained significantly elevated for 5 minutes and returned to baseline by 10 minutes. Our study found longer dwell time of artifical tears when compared with most previous studies.
Yokoi et al.8 used reflective meniscometry to demonstrate a marked increase in lower tear meniscus curvature with the instillation of 1% carboxymethylcellulose. This marked increase in curvature quickly returned to baseline and seemed to stabilize at 5 minutes after instillation. Wang et al.17 also used real-time OCT to study the dynamic tear distribution of 1% carboxymethylcellulose artificial tears. They found that although lower tear meniscus height stayed elevated for 5 minutes, the area had returned to baseline.17 In a follow-up OCT study using the same midviscosity 1% carboxymethylcellulose artificial tears, Palakuru et al.19 found that there was a similar large increase in lower tear meniscus values that returned to baseline by 5 minutes with normal blinking. They hypothesized that the increase in tear meniscus volume was compensated by an increase in output resulting from blinking.19
In a subsequent study, Wang et al.22 used FD-OCT to assess the effect of both 0.5% and 1% carboxymethylcellulose artificial tears. They found that 0.5% carboxymethylcellulose drops elevated both tear meniscus height and area for 5 minutes, whereas 1% carboxymethylcellulose elevated these parameters for 30 minutes. In our study, we used a lower viscosity 0.5% carboxymethylcellulose artificial tear. The instillation of this tear substitute resulted in a similar large increase in lower tear meniscus followed by a rapid decrease over the next 5 minutes. Both our current study and the latest study by Wang et al.22 used FD-OCT to assess tear meniscus dynamics. Both FD-OCT studies found significant elevation of the tear meniscus at 5 minutes. This longer artificial tear-dwell time may reflect the improved sensitivity of higher resolution FD-OCT. With an axial resolution of 5 μm, FD-OCT may provide better visualization of tear meniscus tails and thus allow more accurate and sensitive documentation of tear dynamics. Although artificial tear instillation significantly increased tear meniscus parameters for only 5 minutes, the patients reported improvement in symptoms for much longer.
It is likely that the beneficial effects of artificial tear instillation cannot be solely quantified by changes in tear meniscus dimensions. Other beneficial effects, such as reduction in tear film osmolarity and dilution of inflammatory mediators, may also account for the improvement in symptoms.23–26 Laboratory tests have confirmed that tear hyperosmolarity leads to a cycle of inflammation and damage to the ocular surface. By diluting the osmolarity, artificial tears may decrease epithelial stress, inflammation, and symptoms of irritation.27,28
There are several limitations to our study. Our dry eye population was a heterogenous group of patients, some of whom had severe aqueous tear deficiency as evidenced by a Schirmer test of less than 5 seconds, whereas others had poor quality tears and reflexive tearing. This inclusive population is reflective of the heterogenous nature of a tertiary population of patients with dry eye and thus may show variability in tear meniscus parameters. Two patients had lower punctal plugs, one patient had the lower puncta cauterized, and a fourth patient had the upper puncta cauterized. These confounding variables may have affected the tear meniscus dynamics. A future study with larger numbers could assess the effect of punctal occlusion on tear meniscus parameters. Three patients were excluded due to significant conjunctivochalasis. This is a relatively common finding in the older dry eye population. Mainstone et al.2 found that 7 of 30 subjects had folds of the inferior conjunctiva with irregular tear menisci. Although our images were of sufficient resolution to measure a cross-sectional area in these patients, our triangle approximation computer caliper measurement would not have been accurate.
FD-OCT is able to capture tear meniscus baseline values and tear dynamics after instillation of artificial tears. Artificial tear (0.5% carboxymethylcellulose) instillation dramatically increases tear meniscus initially, and then decays back to baseline values after approximately 5 minutes. FD-OCT may be useful in objectively quantifying the dynamic efficacy of various dry eye treatments ranging from tear supplementation to anti-inflammatories to punctal plugs.