CLINICAL SCIENCE
Novel Tear Interferometer Made of Paper for Lipid Layer Evaluation Ho Sik Hwang, MD, PhD,* Eun Chul Kim, MD, PhD,† and Man Soo Kim, MD, PhD‡
Purpose: To introduce a novel interferometer and provide clinical data based on its use.
Methods: Twenty-two normal subjects, 27 patients with dry eye syndrome, 1 patient with obstructive meibomian gland dysfunction, and 2 patients with ocular graft-versus-host disease were included in the study. We cut a piece of copy paper to create an interferometer comprising a handle, a window, and an illumination area. The patient sits in front of a biomicroscope with his or her head fixed on the headrest. The examiner holds the handle of the interferometer 10 mm from the front of the patient’s eye. The examiner positions the illumination beam of the biomicroscope at the “illumination area.” The examiner then observes a lipid layer interference pattern on the cornea through the window. Results: For the normal subjects, wave or amorphous patterns with white or grey color were observed in interferometry. The mean estimated lipid layer thickness in the dry eye group (43 6 9 nm) was significantly thinner than that of the normal group (59 6 29 nm) (P = 0.021). For the patient with obstructive meibomian gland dysfunction, interferometry revealed a white interference pattern (30 nm). For ocular graft-versus-host disease, the lipid pattern was brown and blue (165 nm). Patients felt no discomfort, and no complications occurred during examination.
Conclusions: We could successfully obtain photographs of the lipid layer pattern with this novel interferometer without additional equipment or cost. This interferometer may be useful for research regarding the pathophysiology, diagnosis, and treatment of dry eye. Key Words: tear film, lipid layer, interferometry, tearscope, dry eye, meibomian gland dysfunction, graft-versus-host disease, LipiView (Cornea 2014;33:826–831)
Received for publication January 9, 2014; revision received April 8, 2014; accepted April 16, 2014. Published online ahead of print June 9, 2014. From the *Department of Ophthalmology, Chuncheon Sacred Heart Hospital College of Medicine, Hallym University, Chuncheon, Korea; †Department of Ophthalmology, Bucheon St Mary’s Hospital College of Medicine, The Catholic University of Korea, Bucheon, Korea; and ‡Department of Ophthalmology, Seoul St Mary’s Hospital, Seoul, Korea. The authors are applying for a patent for this tearscope. The authors have no other conflicts of interest to disclose. Supported by Hallym University Research Fund (HURF-2013-27). Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.corneajrnl.com). Reprints: Man Soo Kim, MD, PhD, Department of Ophthalmology, Seoul St Mary’s Hospital, #505 Banpo-dong, Seocho-gu, Seoul 137-040, Korea (e-mail: [email protected]). Copyright © 2014 by Lippincott Williams & Wilkins
826
| www.corneajrnl.com
reocular tear film comprises mucous, aqueous, and lipid layers. The lipid layer prevents evaporation of the aqueous layer, and is thus very important in dry eye syndrome. Tear interferometry is a noninvasive method for visualizing and evaluating the tear lipid layer. Tear interferometry allows for observation of the behavior of the tear film lipid layer and estimation of its thickness based on the interference colors.1 Tear film dynamics during blinking can also be observed. Because it is not necessary to instill any dye into eyes for interferometry, the lipid layer can be observed in its natural state, without disturbance. Instruments used for this purpose include the Tearscope Plus (Keeler Ophthalmic Instruments, Broomall, PA) and the Kowa DR-1 (Kowa Optimed Inc, Torrance, CA). Unfortunately, however, the diagnostic procedures for interferometry are not widely used in clinical practice because they are often not commercially available. The recently developed LipiView interferometer (TearScience Inc, Morrisville, NC) can measure the lipid layer’s thickness between blinks, and it provides a quantitative assessment in interferometric color units (ICU).2 This instrument is costly, however, and therefore not readily available for clinical practice. We developed a new tear interferometer using a biomicroscope and common copy paper without additional equipment or cost. Here, we introduce this interferometer and provide clinical data based on its use.
P
PATIENTS AND METHODS The study followed the principles of the Declaration of Helsinki, and the Institutional Review Board of Seoul St Mary’s Hospital approved the study. This study was a retrospective medical record review. Twenty-two normal subjects, 27 patients with dry eye syndrome, 1 patient with obstructive meibomian gland dysfunction (MGD), 2 patients with ocular graft-versus-host disease (GVHD), and 3 patients with contamination (face cream and eye ointment) were included in the study. The inclusion criteria of dry eye syndrome were typical dry eye symptoms (ie, dryness, foreign body sensation, ocular fatigue, burning, stinging, itching, soreness, heaviness of eyelids, photophobia) and the tear break-up time (BUT) of less than 10 seconds or corneal and conjunctival surface staining of higher than grade 1 according to the Oxford scoring scheme.3 The obstructive MGD diagnosis criteria were meibomian gland drop out in infrared meibography, terminal duct obstruction, plugging of the meibomian glands, turbid secretions or turbid secretions with clumps, inflammation and swelling of the eyelid margin, vascular Cornea Volume 33, Number 8, August 2014
Cornea Volume 33, Number 8, August 2014
injection of the posterior lid margin, and symptoms of MGD for at least 3 months.4 The new interferometer used the following techniques. For the interferometer, we cut a piece of common copy paper (weight 75 g/m2) to create a rectangle 70 mm wide and 20 mm high and rounded one end. We punched a 5 mm-sized circular hole at the rounded end as an observation window. Thus, the interferometer comprised a handle, a window, and an illumination area (Fig. 1A). The patient sits in front of a biomicroscope with his or her head fixed on the headrest. The examiner holds the handle of the interferometer 10 mm from the front of the patient’s eye (Fig. 1B). The window of the interferometer is positioned at the center of the cornea. The examiner positions the illumination beam of the biomicroscope at the “illumination area.” The examiner then observes the reflected image of the “illumination area” on the cornea through the window under low magnification (·10, objective). The examiner then increases the magnification to ·25. By moving the biomicroscope forward and backward, the examiner can observe a clear image of the reflection of the paper’s texture from the “illumination area.” The examiner slowly pulls the biomicroscope away from the patient until a lipid layer interference pattern appears (see Video, Supplemental Digital Content 1, http://links.lww.com/ICO/A225). We obtained photographs of the lipid layer pattern or saved them as movies. The photographs were captured by the charge-coupled device camera. The light intensity was adjusted by changing the biomicroscope intensity and the aperture of the iris of a charged-coupled device camera to avoid overexposure or underexposure. The lipid layer interference patterns were graded according to Korb’s classification of lipid layer interference patterns as follows: (1) white (estimated lipid layer thickness 30 nm), (2) grey/white (45 nm), (3) grey (60 nm), (4) grey/ yellow (75 nm), (5) yellow (90 nm), (6) yellow/brown (105 nm), (7) brown/yellow (120 nm), (8) brown (135 nm), (9) brown/blue (150 nm), (10) blue/brown (165 nm), and (11) blue (180 nm).5 We never touched the eyelids before or during interferometry, and thus avoided artifacts from lid manipulation. All interferometries were performed by a single ophthalmologist (H.S.H.). After the interferometry, we measured the BUT, obtained a fluorescein corneal stain, lid margin photograph, and infrared meibography in sequences. Grading of the corneal and conjunctival surface staining was conducted according to the Oxford scoring scheme.3 We used the same procedures for infrared meibography as used in previous studies.6,7 In infrared meibography, meibomian gland loss was graded according to Arita’s criteria.8 Statistical analyses for comparison between dry eye (27 right eyes of 27 patients) and normal groups (22 right eyes of 22 subjects) were performed using the Mann–Whitney test and Fisher exact test in SPSS version 12.0 (SPSS Inc, Chicago, IL). Results were considered statistically significant if the P value was ,0.05.
RESULTS Subject A was a healthy 30-year-old woman. In interferometry, grey and yellow wave patterns (grey/yellow, estimated 2014 Lippincott Williams & Wilkins
Novel Tear Interferometer Made of Paper
FIGURE 1. The design of a new interferometer and the way of use. For the interferometer, we cut a piece of common copy paper to create a rectangle 70 mm wide and 20 mm high and rounded one end. We punched a 5-mm-sized circular hole at the rounded end as an observation window. A, Thus, the interferometer comprised a handle, a window, and an illumination area. For interferometry, the patient sits in front of a biomicroscope with his head fixed on the headrest. B, The examiner holds the handle of the interferometer 10 mm from the front of the patient’s eye.
thickness 75 nm), with black background were observed (Fig. 2). We observed no clumps on the tear film. In movies, immediately after blinking, the thin (dark) lipid layer became thicker (light) and the wave-pattern was reformed (see Video, Supplemental Digital Content 2, http://links.lww.com/ICO/A226). For the other normal subjects B–E, white, white/grey, or grey colors were observed in interferometry (Fig. 2). Patient F was a 50-year-old woman who complained of dryness in both eyes (Fig. 3). In interferometry, we observed white/grey color (estimated thickness 45 nm). The BUT was 9 seconds, and diffuse superficial punctuate keratitis at the interior cornea was observed. The lid margin was normal. In infrared meibography, a grade 0 meiboscore was observed. The patient was diagnosed with moderate dry eye syndrome (aqueous deficiency type). Patient G was another patient with dry eye syndrome (see Video, Supplemental Digital Content 3, www.corneajrnl.com |
827
Hwang et al
Cornea Volume 33, Number 8, August 2014
FIGURE 2. The interference patterns of normal subjects. Subject A was a healthy 30-year-old woman. In interferometry, grey and yellow color (estimated thickness 75 nm), with black background were observed. For the other normal subjects (subject B–E), white, white/ grey, or grey colors were observed in interferometry.
http://links.lww.com/ICO/A227). In interferometry, the lipid layer appeared very dark and very thin. The lipid layer was not even, but fragmented film plaque was floating. Table 1 shows the comparison of clinical characteristics between the dry eye and normal group. The mean estimated lipid layer thickness in the dry eye group (43 6 9 nm) was significantly thinner than that of the normal group (59 6 29 nm) (P = 0.021). The Oxford score of the dry eye group (1.1 6 1.1) was significantly higher than that of the normal group (0.0 6 0.0) (P = 0.000). The lower
meibomian glands of the dry eye group (1.2 6 1.0) had significantly higher meiboscore than that of the normal group (0.6 6 0.6). But, the BUT and the meiboscore of the upper meibomian glands did not show significant differences between the 2 groups. Patient H was a 73-year-old man who complained of foreign body sensation and ocular pain in both eyes (Fig. 4). Interferometry revealed a white pattern (white, estimated thickness 30 nm). The lid margin showed anteroplacement of the mucocutaneous junction and obliteration of meibomian gland
FIGURE 3. Patient F was a 50-yearold woman who complained of dryness in both eyes. A, In interferometry, we observed white/grey color (estimated thickness 45 nm). B, The BUT was 9 seconds, and diffuse superficial punctuate keratitis at interior cornea was observed. C, The lid margin was normal. D, In infrared meibography, a grade 0 meiboscore was observed. The patient was diagnosed with moderate dry eye syndrome (aqueous deficiency type).
828
| www.corneajrnl.com
2014 Lippincott Williams & Wilkins
Cornea Volume 33, Number 8, August 2014
Novel Tear Interferometer Made of Paper
TABLE 1. Comparison of Clinical Characteristics Between the Dry Eye Group and the Normal Group Parameters Age, yr Sex, n (%) Male Females Lipid layer thickness, nm BUT, s Oxford score Meiboscores Upper meibomian glands Lower meibomian glands
Dry Eye Group (n = 27)
Normal Group (n = 22)
P
54.2 6 12.6
52.9 6 17.4
0.747*
11 (41) 16 (59) 43 6 9 5.2 6 2.9 1.1 6 1.1
8 (36) 14 (64) 59 6 29 6.7 6 5.1 0.0 6 0.0
1.000†
1.0 6 0.8 1.2 6 1.0
0.7 6 0.7 0.6 6 0.6
0.021* 0.293* 0.000* 0.191* 0.033*
*Mann–Whitney test. †Fisher’s exact test.
orifices. In infrared meibography, we observed no normal meibomian glands (meiboscore grade 3). He was diagnosed with obstructive MGD. Patient I was a 65-year-old patient diagnosed with ocular GVHD, after bone marrow transplantation (Fig. 5). In interferometry, the lipid pattern was brown and blue (brown/blue, estimated thickness 165 nm). We observed larger clumps. The BUT was 2 seconds, and diffuse superficial punctuate keratitis at the inferior cornea was observed. The lid margin was covered by a turbid secretion. In infrared meibography, the meiboscore was grade 2. Patient J was another patient with ocular GVHD. In videos, the lipid layer was very thin, and we observed the rupture of small oil drops (see Video, Supplemental Digital Content 4, http://links.lww.com/ICO/A228).
The lipid layer contaminated with a face cream in patient K looked like marbling, and the lipid layer pattern contaminated with an eye ointment in patient L was very coarse and colorful (not shown here). Patient M had used an eye ointment after keratoplasty (see Video, Supplemental Digital Content 5, http://links.lww.com/ICO/A229). Examination time for interferometry was less than 5 seconds. Patients felt no discomfort and no complications occurred during examination.
DISCUSSION We report the development of a new interferometer and the ability to conveniently obtain photographs of the lipid layer using this new device. The interferometer uses “thin film interference phenomena” like other tear interferometers. When the illumination beam of the biomicroscope is projected at the “illumination area,” it serves as a new light source. The scattered light from the “illumination area” is reflected at the corneal surface and reaches the examiner’s eyes. When the light is reflected at the corneal surface, a portion of it is reflected at the air–lipid layer interface and the other portion of it penetrates the lipid layer and is reflected at the lipid layer–aqueous layer interface. The 2 reflected lights interfere with each other and show thin film interference phenomena. The interference pattern and color depend on the lipid layer thickness and refractive index of the lipid. This interferometer has several advantages. First, photographs of the lipid layer pattern can be obtained without additional equipment or cost. Only common copy paper is necessary. Therefore, every ophthalmologist can evaluate the lipid layer of tear film using this technique. The image quality is not inferior to that of other, expensive, interferometers.
FIGURE 4. Patient H was a 73-yearold man who complained of foreign body sensation and ocular pain in both eyes. A, Interferometry revealed a white pattern (estimated thickness 30 nm). B, The lid margin showed anteroplacement of the mucocutaneous junction and obliteration of meibomian gland orifices. C, In infrared meibography, we observed no normal meibomian glands (meiboscore grade 3). He was diagnosed with obstructive MGD. 2014 Lippincott Williams & Wilkins
www.corneajrnl.com |
829
Hwang et al
Cornea Volume 33, Number 8, August 2014
FIGURE 5. Patient I was a 65-yearold patient diagnosed with ocular GVHD, after bone marrow transplantation. A, In interferometry, the lipid pattern was brown and blue (brown/blue, estimated thickness 165 nm). We observed larger clumps. B, The BUT was 2 seconds, and diffuse superficial punctuate keratitis at the inferior cornea was observed. C, The lid margin was covered by a turbid secretion. D, In infrared meibography, the meiboscore was grade 2.
Second, this interferometer is easy to use, and the examination time was less than 5 seconds. Third, a photograph can be obtained without disturbing the lipid layer or touching the eyelids, and the results are reliable. If the patient’s eyelids require any manipulation, meibum is expressed from the lid margin. This is different from the natural conditions of the ocular surface. Fourth, the dark background in this technique highlights the lipid layer patterns. Other interferometers, such as the Tearscope Plus, do not have a dark background. Fifth, additional illumination is not necessary. Only the illumination from the biomicroscope and rather weak light are needed, which ensures that this interferometer is safe. This interferometer has some disadvantages. First, although is not difficult to use, some training is necessary. Second, the entire cornea cannot be observed, only the local area. Third, this interferometer is not attached to the biomicroscope, and the left hand is needed to hold the interferometer. Fourth, the thickness of the lipid layer cannot be quantitatively measured, but this drawback is similar to that of other interferometers, except for the LipiView. Semiquantitative measurements, however, can be obtained. In this research, the lipid layer interference patterns were graded according to Korb’s classification. There are other classification systems similar to it.1,9–11 The Korb’s classification system has some advantages compared with the others. Because the system classifies the interference patterns according to only their dominant color, consistency between interpreters is higher than those of the other classification systems. Also, it gives estimated thickness of lipid layer, therefore it can be used for quantitative analysis. In the case with obstructive MGD (patient H), the estimated lipid layer thickness was only 30 nm, because the oil secretion from meibomian glands was decreased. In
830
| www.corneajrnl.com
the case with ocular GVHD (patient I), the lipid pattern was brown and blue (brown/blue, estimated thickness 165 nm). This is consistent with other studies regarding lipid layer thickness in ocular GVHD.12–14 Goto and Tseng12,13 reported that the thick lipid layer in eyes with aqueous tear deficiency results from the retardation of lipid spread, which leads to uneven distribution of the lipid film. In Table 1, the BUT and meiboscores of the upper lids were not significantly different between the dry eye and normal groups, unexpectedly. Obstructive MGD is thought to be one of the major causes of dry eye syndrome. Theoretically, in obstructive MGD, lid margin abnormality such as meibomian gland plugging, meibomian gland drop out in infrared meibography, thin lipid layer, short BUT, and high scores of corneal and conjunctival stain must be present. However, it is not always correct in actual patients. For example, in our pilot study, BUT was poorly correlated with meiboscores of the upper and lower lids (not shown here). We are analyzing more clinical data to investigate these relationships between meiboscores, lid margin abnormality, lipid layer thickness, BUT, corneal and conjunctival stain in normal, dry eye syndrome, MGD, GVHD for our next study. Evaluation of the tear film is very important in the diagnosis and treatment of dry eye syndrome. It is not widely performed, however, because interferometers are often not commercially available or are too expensive for most ophthalmologists. Here, we describe a technique for taking photographs of the lipid layer of tear film using common copy paper. No expensive equipment is necessary for performing this technique. Because the lipid layer interference pattern changes during blinking, this method is also very useful for making videos. This interferometer will be useful for research regarding the pathophysiology, diagnosis, and treatment of dry eye. 2014 Lippincott Williams & Wilkins
Cornea Volume 33, Number 8, August 2014
REFERENCES 1. Yokoi N, Takehisa Y, Kinoshita S. Correlation of tear lipid layer interference patterns with the diagnosis and severity of dry eye. Am J Ophthalmol. 1996;122:818–824. 2. Eom Y, Lee JS, Kang SY, et al. Correlation between quantitative measurements of tear film lipid layer thickness and meibomian gland loss in patients with obstructive meibomian gland dysfunction and normal controls. Am J Ophthalmol. 2013;155:1104–1110.e2. 3. Bron AJ, Evans VE, Smith JA. Grading of corneal and conjunctival staining in the context of other dry eye tests. Cornea. 2003;22: 640–650. 4. Foulks GN, Bron AJ. Meibomian gland dysfunction: a clinical scheme for description, diagnosis, classification, and grading. Ocul Surf. 2003;1: 107–126. 5. Korb DR, Baron DF, Herman JP, et al. Tear film lipid layer thickness as a function of blinking. Cornea. 1994;13:354–359. 6. Hwang HS, Park CW, Joo CK. Novel noncontact meibography with anterior segment optical coherence tomography: Hosik meibography. Cornea. 2013;32:40–43.
2014 Lippincott Williams & Wilkins
Novel Tear Interferometer Made of Paper
7. Hwang HS, Shin JG, Lee BH, et al. In vivo 3D meibography of the human eyelid using real time imaging Fourier-domain OCT. PLoS One. 2013;8:e67143. 8. Arita R, Itoh K, Inoue K, et al. Noncontact infrared meibography to document age-related changes of the meibomian glands in a normal population. Ophthalmology. 2008;115:911–915. 9. Guillon JP. Non-invasive Tearscope Plus routine for contact lens fitting. Cont Lens Anterior Eye. 1998;21(suppl 1):S31–S40. 10. Isenberg SJ, Del Signore M, Chen A, et al. The lipid layer and stability of the preocular tear film in newborns and infants. Ophthalmology. 2003; 110:1408–1411. 11. Maïssa C, Guillon M. Tear film dynamics and lipid layer characteristics— effect of age and gender. Cont Lens Anterior Eye. 2010;33:176–182. 12. Goto E, Tseng SC. Differentiation of lipid tear deficiency dry eye by kinetic analysis of tear interference images. Arch Ophthalmol. 2003;121:173–180. 13. Goto E. Quantification of tear interference image: tear fluid surface nanotechnology. Cornea. 2004;23:S20–S24. 14. Ban Y, Ogawa Y, Goto E, et al. Tear function and lipid layer alterations in dry eye patients with chronic graft-vs-host disease. Eye (Lond). 2009;23:202–208.
www.corneajrnl.com |
831
Novel Tear Interferometer Made of Paper for Lipid Layer Evaluation Ho Sik Hwang, MD, PhD,* Eun Chul Kim, MD, PhD,† and Man Soo Kim, MD, PhD‡
Purpose: To introduce a novel interferometer and provide clinical data based on its use.
Methods: Twenty-two normal subjects, 27 patients with dry eye syndrome, 1 patient with obstructive meibomian gland dysfunction, and 2 patients with ocular graft-versus-host disease were included in the study. We cut a piece of copy paper to create an interferometer comprising a handle, a window, and an illumination area. The patient sits in front of a biomicroscope with his or her head fixed on the headrest. The examiner holds the handle of the interferometer 10 mm from the front of the patient’s eye. The examiner positions the illumination beam of the biomicroscope at the “illumination area.” The examiner then observes a lipid layer interference pattern on the cornea through the window. Results: For the normal subjects, wave or amorphous patterns with white or grey color were observed in interferometry. The mean estimated lipid layer thickness in the dry eye group (43 6 9 nm) was significantly thinner than that of the normal group (59 6 29 nm) (P = 0.021). For the patient with obstructive meibomian gland dysfunction, interferometry revealed a white interference pattern (30 nm). For ocular graft-versus-host disease, the lipid pattern was brown and blue (165 nm). Patients felt no discomfort, and no complications occurred during examination.
Conclusions: We could successfully obtain photographs of the lipid layer pattern with this novel interferometer without additional equipment or cost. This interferometer may be useful for research regarding the pathophysiology, diagnosis, and treatment of dry eye. Key Words: tear film, lipid layer, interferometry, tearscope, dry eye, meibomian gland dysfunction, graft-versus-host disease, LipiView (Cornea 2014;33:826–831)
Received for publication January 9, 2014; revision received April 8, 2014; accepted April 16, 2014. Published online ahead of print June 9, 2014. From the *Department of Ophthalmology, Chuncheon Sacred Heart Hospital College of Medicine, Hallym University, Chuncheon, Korea; †Department of Ophthalmology, Bucheon St Mary’s Hospital College of Medicine, The Catholic University of Korea, Bucheon, Korea; and ‡Department of Ophthalmology, Seoul St Mary’s Hospital, Seoul, Korea. The authors are applying for a patent for this tearscope. The authors have no other conflicts of interest to disclose. Supported by Hallym University Research Fund (HURF-2013-27). Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.corneajrnl.com). Reprints: Man Soo Kim, MD, PhD, Department of Ophthalmology, Seoul St Mary’s Hospital, #505 Banpo-dong, Seocho-gu, Seoul 137-040, Korea (e-mail: [email protected]). Copyright © 2014 by Lippincott Williams & Wilkins
826
| www.corneajrnl.com
reocular tear film comprises mucous, aqueous, and lipid layers. The lipid layer prevents evaporation of the aqueous layer, and is thus very important in dry eye syndrome. Tear interferometry is a noninvasive method for visualizing and evaluating the tear lipid layer. Tear interferometry allows for observation of the behavior of the tear film lipid layer and estimation of its thickness based on the interference colors.1 Tear film dynamics during blinking can also be observed. Because it is not necessary to instill any dye into eyes for interferometry, the lipid layer can be observed in its natural state, without disturbance. Instruments used for this purpose include the Tearscope Plus (Keeler Ophthalmic Instruments, Broomall, PA) and the Kowa DR-1 (Kowa Optimed Inc, Torrance, CA). Unfortunately, however, the diagnostic procedures for interferometry are not widely used in clinical practice because they are often not commercially available. The recently developed LipiView interferometer (TearScience Inc, Morrisville, NC) can measure the lipid layer’s thickness between blinks, and it provides a quantitative assessment in interferometric color units (ICU).2 This instrument is costly, however, and therefore not readily available for clinical practice. We developed a new tear interferometer using a biomicroscope and common copy paper without additional equipment or cost. Here, we introduce this interferometer and provide clinical data based on its use.
P
PATIENTS AND METHODS The study followed the principles of the Declaration of Helsinki, and the Institutional Review Board of Seoul St Mary’s Hospital approved the study. This study was a retrospective medical record review. Twenty-two normal subjects, 27 patients with dry eye syndrome, 1 patient with obstructive meibomian gland dysfunction (MGD), 2 patients with ocular graft-versus-host disease (GVHD), and 3 patients with contamination (face cream and eye ointment) were included in the study. The inclusion criteria of dry eye syndrome were typical dry eye symptoms (ie, dryness, foreign body sensation, ocular fatigue, burning, stinging, itching, soreness, heaviness of eyelids, photophobia) and the tear break-up time (BUT) of less than 10 seconds or corneal and conjunctival surface staining of higher than grade 1 according to the Oxford scoring scheme.3 The obstructive MGD diagnosis criteria were meibomian gland drop out in infrared meibography, terminal duct obstruction, plugging of the meibomian glands, turbid secretions or turbid secretions with clumps, inflammation and swelling of the eyelid margin, vascular Cornea Volume 33, Number 8, August 2014
Cornea Volume 33, Number 8, August 2014
injection of the posterior lid margin, and symptoms of MGD for at least 3 months.4 The new interferometer used the following techniques. For the interferometer, we cut a piece of common copy paper (weight 75 g/m2) to create a rectangle 70 mm wide and 20 mm high and rounded one end. We punched a 5 mm-sized circular hole at the rounded end as an observation window. Thus, the interferometer comprised a handle, a window, and an illumination area (Fig. 1A). The patient sits in front of a biomicroscope with his or her head fixed on the headrest. The examiner holds the handle of the interferometer 10 mm from the front of the patient’s eye (Fig. 1B). The window of the interferometer is positioned at the center of the cornea. The examiner positions the illumination beam of the biomicroscope at the “illumination area.” The examiner then observes the reflected image of the “illumination area” on the cornea through the window under low magnification (·10, objective). The examiner then increases the magnification to ·25. By moving the biomicroscope forward and backward, the examiner can observe a clear image of the reflection of the paper’s texture from the “illumination area.” The examiner slowly pulls the biomicroscope away from the patient until a lipid layer interference pattern appears (see Video, Supplemental Digital Content 1, http://links.lww.com/ICO/A225). We obtained photographs of the lipid layer pattern or saved them as movies. The photographs were captured by the charge-coupled device camera. The light intensity was adjusted by changing the biomicroscope intensity and the aperture of the iris of a charged-coupled device camera to avoid overexposure or underexposure. The lipid layer interference patterns were graded according to Korb’s classification of lipid layer interference patterns as follows: (1) white (estimated lipid layer thickness 30 nm), (2) grey/white (45 nm), (3) grey (60 nm), (4) grey/ yellow (75 nm), (5) yellow (90 nm), (6) yellow/brown (105 nm), (7) brown/yellow (120 nm), (8) brown (135 nm), (9) brown/blue (150 nm), (10) blue/brown (165 nm), and (11) blue (180 nm).5 We never touched the eyelids before or during interferometry, and thus avoided artifacts from lid manipulation. All interferometries were performed by a single ophthalmologist (H.S.H.). After the interferometry, we measured the BUT, obtained a fluorescein corneal stain, lid margin photograph, and infrared meibography in sequences. Grading of the corneal and conjunctival surface staining was conducted according to the Oxford scoring scheme.3 We used the same procedures for infrared meibography as used in previous studies.6,7 In infrared meibography, meibomian gland loss was graded according to Arita’s criteria.8 Statistical analyses for comparison between dry eye (27 right eyes of 27 patients) and normal groups (22 right eyes of 22 subjects) were performed using the Mann–Whitney test and Fisher exact test in SPSS version 12.0 (SPSS Inc, Chicago, IL). Results were considered statistically significant if the P value was ,0.05.
RESULTS Subject A was a healthy 30-year-old woman. In interferometry, grey and yellow wave patterns (grey/yellow, estimated 2014 Lippincott Williams & Wilkins
Novel Tear Interferometer Made of Paper
FIGURE 1. The design of a new interferometer and the way of use. For the interferometer, we cut a piece of common copy paper to create a rectangle 70 mm wide and 20 mm high and rounded one end. We punched a 5-mm-sized circular hole at the rounded end as an observation window. A, Thus, the interferometer comprised a handle, a window, and an illumination area. For interferometry, the patient sits in front of a biomicroscope with his head fixed on the headrest. B, The examiner holds the handle of the interferometer 10 mm from the front of the patient’s eye.
thickness 75 nm), with black background were observed (Fig. 2). We observed no clumps on the tear film. In movies, immediately after blinking, the thin (dark) lipid layer became thicker (light) and the wave-pattern was reformed (see Video, Supplemental Digital Content 2, http://links.lww.com/ICO/A226). For the other normal subjects B–E, white, white/grey, or grey colors were observed in interferometry (Fig. 2). Patient F was a 50-year-old woman who complained of dryness in both eyes (Fig. 3). In interferometry, we observed white/grey color (estimated thickness 45 nm). The BUT was 9 seconds, and diffuse superficial punctuate keratitis at the interior cornea was observed. The lid margin was normal. In infrared meibography, a grade 0 meiboscore was observed. The patient was diagnosed with moderate dry eye syndrome (aqueous deficiency type). Patient G was another patient with dry eye syndrome (see Video, Supplemental Digital Content 3, www.corneajrnl.com |
827
Hwang et al
Cornea Volume 33, Number 8, August 2014
FIGURE 2. The interference patterns of normal subjects. Subject A was a healthy 30-year-old woman. In interferometry, grey and yellow color (estimated thickness 75 nm), with black background were observed. For the other normal subjects (subject B–E), white, white/ grey, or grey colors were observed in interferometry.
http://links.lww.com/ICO/A227). In interferometry, the lipid layer appeared very dark and very thin. The lipid layer was not even, but fragmented film plaque was floating. Table 1 shows the comparison of clinical characteristics between the dry eye and normal group. The mean estimated lipid layer thickness in the dry eye group (43 6 9 nm) was significantly thinner than that of the normal group (59 6 29 nm) (P = 0.021). The Oxford score of the dry eye group (1.1 6 1.1) was significantly higher than that of the normal group (0.0 6 0.0) (P = 0.000). The lower
meibomian glands of the dry eye group (1.2 6 1.0) had significantly higher meiboscore than that of the normal group (0.6 6 0.6). But, the BUT and the meiboscore of the upper meibomian glands did not show significant differences between the 2 groups. Patient H was a 73-year-old man who complained of foreign body sensation and ocular pain in both eyes (Fig. 4). Interferometry revealed a white pattern (white, estimated thickness 30 nm). The lid margin showed anteroplacement of the mucocutaneous junction and obliteration of meibomian gland
FIGURE 3. Patient F was a 50-yearold woman who complained of dryness in both eyes. A, In interferometry, we observed white/grey color (estimated thickness 45 nm). B, The BUT was 9 seconds, and diffuse superficial punctuate keratitis at interior cornea was observed. C, The lid margin was normal. D, In infrared meibography, a grade 0 meiboscore was observed. The patient was diagnosed with moderate dry eye syndrome (aqueous deficiency type).
828
| www.corneajrnl.com
2014 Lippincott Williams & Wilkins
Cornea Volume 33, Number 8, August 2014
Novel Tear Interferometer Made of Paper
TABLE 1. Comparison of Clinical Characteristics Between the Dry Eye Group and the Normal Group Parameters Age, yr Sex, n (%) Male Females Lipid layer thickness, nm BUT, s Oxford score Meiboscores Upper meibomian glands Lower meibomian glands
Dry Eye Group (n = 27)
Normal Group (n = 22)
P
54.2 6 12.6
52.9 6 17.4
0.747*
11 (41) 16 (59) 43 6 9 5.2 6 2.9 1.1 6 1.1
8 (36) 14 (64) 59 6 29 6.7 6 5.1 0.0 6 0.0
1.000†
1.0 6 0.8 1.2 6 1.0
0.7 6 0.7 0.6 6 0.6
0.021* 0.293* 0.000* 0.191* 0.033*
*Mann–Whitney test. †Fisher’s exact test.
orifices. In infrared meibography, we observed no normal meibomian glands (meiboscore grade 3). He was diagnosed with obstructive MGD. Patient I was a 65-year-old patient diagnosed with ocular GVHD, after bone marrow transplantation (Fig. 5). In interferometry, the lipid pattern was brown and blue (brown/blue, estimated thickness 165 nm). We observed larger clumps. The BUT was 2 seconds, and diffuse superficial punctuate keratitis at the inferior cornea was observed. The lid margin was covered by a turbid secretion. In infrared meibography, the meiboscore was grade 2. Patient J was another patient with ocular GVHD. In videos, the lipid layer was very thin, and we observed the rupture of small oil drops (see Video, Supplemental Digital Content 4, http://links.lww.com/ICO/A228).
The lipid layer contaminated with a face cream in patient K looked like marbling, and the lipid layer pattern contaminated with an eye ointment in patient L was very coarse and colorful (not shown here). Patient M had used an eye ointment after keratoplasty (see Video, Supplemental Digital Content 5, http://links.lww.com/ICO/A229). Examination time for interferometry was less than 5 seconds. Patients felt no discomfort and no complications occurred during examination.
DISCUSSION We report the development of a new interferometer and the ability to conveniently obtain photographs of the lipid layer using this new device. The interferometer uses “thin film interference phenomena” like other tear interferometers. When the illumination beam of the biomicroscope is projected at the “illumination area,” it serves as a new light source. The scattered light from the “illumination area” is reflected at the corneal surface and reaches the examiner’s eyes. When the light is reflected at the corneal surface, a portion of it is reflected at the air–lipid layer interface and the other portion of it penetrates the lipid layer and is reflected at the lipid layer–aqueous layer interface. The 2 reflected lights interfere with each other and show thin film interference phenomena. The interference pattern and color depend on the lipid layer thickness and refractive index of the lipid. This interferometer has several advantages. First, photographs of the lipid layer pattern can be obtained without additional equipment or cost. Only common copy paper is necessary. Therefore, every ophthalmologist can evaluate the lipid layer of tear film using this technique. The image quality is not inferior to that of other, expensive, interferometers.
FIGURE 4. Patient H was a 73-yearold man who complained of foreign body sensation and ocular pain in both eyes. A, Interferometry revealed a white pattern (estimated thickness 30 nm). B, The lid margin showed anteroplacement of the mucocutaneous junction and obliteration of meibomian gland orifices. C, In infrared meibography, we observed no normal meibomian glands (meiboscore grade 3). He was diagnosed with obstructive MGD. 2014 Lippincott Williams & Wilkins
www.corneajrnl.com |
829
Hwang et al
Cornea Volume 33, Number 8, August 2014
FIGURE 5. Patient I was a 65-yearold patient diagnosed with ocular GVHD, after bone marrow transplantation. A, In interferometry, the lipid pattern was brown and blue (brown/blue, estimated thickness 165 nm). We observed larger clumps. B, The BUT was 2 seconds, and diffuse superficial punctuate keratitis at the inferior cornea was observed. C, The lid margin was covered by a turbid secretion. D, In infrared meibography, the meiboscore was grade 2.
Second, this interferometer is easy to use, and the examination time was less than 5 seconds. Third, a photograph can be obtained without disturbing the lipid layer or touching the eyelids, and the results are reliable. If the patient’s eyelids require any manipulation, meibum is expressed from the lid margin. This is different from the natural conditions of the ocular surface. Fourth, the dark background in this technique highlights the lipid layer patterns. Other interferometers, such as the Tearscope Plus, do not have a dark background. Fifth, additional illumination is not necessary. Only the illumination from the biomicroscope and rather weak light are needed, which ensures that this interferometer is safe. This interferometer has some disadvantages. First, although is not difficult to use, some training is necessary. Second, the entire cornea cannot be observed, only the local area. Third, this interferometer is not attached to the biomicroscope, and the left hand is needed to hold the interferometer. Fourth, the thickness of the lipid layer cannot be quantitatively measured, but this drawback is similar to that of other interferometers, except for the LipiView. Semiquantitative measurements, however, can be obtained. In this research, the lipid layer interference patterns were graded according to Korb’s classification. There are other classification systems similar to it.1,9–11 The Korb’s classification system has some advantages compared with the others. Because the system classifies the interference patterns according to only their dominant color, consistency between interpreters is higher than those of the other classification systems. Also, it gives estimated thickness of lipid layer, therefore it can be used for quantitative analysis. In the case with obstructive MGD (patient H), the estimated lipid layer thickness was only 30 nm, because the oil secretion from meibomian glands was decreased. In
830
| www.corneajrnl.com
the case with ocular GVHD (patient I), the lipid pattern was brown and blue (brown/blue, estimated thickness 165 nm). This is consistent with other studies regarding lipid layer thickness in ocular GVHD.12–14 Goto and Tseng12,13 reported that the thick lipid layer in eyes with aqueous tear deficiency results from the retardation of lipid spread, which leads to uneven distribution of the lipid film. In Table 1, the BUT and meiboscores of the upper lids were not significantly different between the dry eye and normal groups, unexpectedly. Obstructive MGD is thought to be one of the major causes of dry eye syndrome. Theoretically, in obstructive MGD, lid margin abnormality such as meibomian gland plugging, meibomian gland drop out in infrared meibography, thin lipid layer, short BUT, and high scores of corneal and conjunctival stain must be present. However, it is not always correct in actual patients. For example, in our pilot study, BUT was poorly correlated with meiboscores of the upper and lower lids (not shown here). We are analyzing more clinical data to investigate these relationships between meiboscores, lid margin abnormality, lipid layer thickness, BUT, corneal and conjunctival stain in normal, dry eye syndrome, MGD, GVHD for our next study. Evaluation of the tear film is very important in the diagnosis and treatment of dry eye syndrome. It is not widely performed, however, because interferometers are often not commercially available or are too expensive for most ophthalmologists. Here, we describe a technique for taking photographs of the lipid layer of tear film using common copy paper. No expensive equipment is necessary for performing this technique. Because the lipid layer interference pattern changes during blinking, this method is also very useful for making videos. This interferometer will be useful for research regarding the pathophysiology, diagnosis, and treatment of dry eye. 2014 Lippincott Williams & Wilkins
Cornea Volume 33, Number 8, August 2014
REFERENCES 1. Yokoi N, Takehisa Y, Kinoshita S. Correlation of tear lipid layer interference patterns with the diagnosis and severity of dry eye. Am J Ophthalmol. 1996;122:818–824. 2. Eom Y, Lee JS, Kang SY, et al. Correlation between quantitative measurements of tear film lipid layer thickness and meibomian gland loss in patients with obstructive meibomian gland dysfunction and normal controls. Am J Ophthalmol. 2013;155:1104–1110.e2. 3. Bron AJ, Evans VE, Smith JA. Grading of corneal and conjunctival staining in the context of other dry eye tests. Cornea. 2003;22: 640–650. 4. Foulks GN, Bron AJ. Meibomian gland dysfunction: a clinical scheme for description, diagnosis, classification, and grading. Ocul Surf. 2003;1: 107–126. 5. Korb DR, Baron DF, Herman JP, et al. Tear film lipid layer thickness as a function of blinking. Cornea. 1994;13:354–359. 6. Hwang HS, Park CW, Joo CK. Novel noncontact meibography with anterior segment optical coherence tomography: Hosik meibography. Cornea. 2013;32:40–43.
2014 Lippincott Williams & Wilkins
Novel Tear Interferometer Made of Paper
7. Hwang HS, Shin JG, Lee BH, et al. In vivo 3D meibography of the human eyelid using real time imaging Fourier-domain OCT. PLoS One. 2013;8:e67143. 8. Arita R, Itoh K, Inoue K, et al. Noncontact infrared meibography to document age-related changes of the meibomian glands in a normal population. Ophthalmology. 2008;115:911–915. 9. Guillon JP. Non-invasive Tearscope Plus routine for contact lens fitting. Cont Lens Anterior Eye. 1998;21(suppl 1):S31–S40. 10. Isenberg SJ, Del Signore M, Chen A, et al. The lipid layer and stability of the preocular tear film in newborns and infants. Ophthalmology. 2003; 110:1408–1411. 11. Maïssa C, Guillon M. Tear film dynamics and lipid layer characteristics— effect of age and gender. Cont Lens Anterior Eye. 2010;33:176–182. 12. Goto E, Tseng SC. Differentiation of lipid tear deficiency dry eye by kinetic analysis of tear interference images. Arch Ophthalmol. 2003;121:173–180. 13. Goto E. Quantification of tear interference image: tear fluid surface nanotechnology. Cornea. 2004;23:S20–S24. 14. Ban Y, Ogawa Y, Goto E, et al. Tear function and lipid layer alterations in dry eye patients with chronic graft-vs-host disease. Eye (Lond). 2009;23:202–208.
www.corneajrnl.com |
831
