The Investigation of Tear Film Osmolality as a Clinical Instrument Used in Assessments of the Tear Film and Dry Eye Disease

Abstract

Introduction: Tear film osmolality is a product of the varying concentrations of dissolved solutes (proteins, lipids and mucins) in the tear fluid. Research suggests that a hyperosmotic tear film is a trait common to all forms of dry eye, and it may be the driving force causing the discomfort, ocular surface damage and inflammation found in both evaporative and tear deficient forms of dry eye disease. Tear film osmolality has been proposed to be the “gold standard” diagnostic test for the evaluation of dry eye disease, as a distinct separation between tear film osmolalities in normal and dry-eyed (aqueous deficient or evaporative) populations has become evident.

Historically, tear film osmolality could only be measured in a laboratory setting and required a highly skilled technician to use the instrumentation. The recent development of easy-to-use, small volume osmometers has made it possible for tear film osmolality to be measured clinically. As these instruments are quite new, there has been very little research completed with them. Therefore, a series of studies was conducted to investigate the utility of one of these new osmometers – the Advanced Instruments Model 3100 Nanolitre Osmometer.

The specific aims of each chapter were: - Chapter 3: To determine if the Advanced Instruments Model 3100 Nanolitre Osmometer was capable of quantitatively measuring tear film osmolality in a normal population, using 0.5μL tear samples. - Chapter 4: Previous studies have shown the Advanced Instruments Model 3100 Nanolitre Osmometer not significantly different from another commercially available osmometer (Wescor Vapor Pressure Osmometer) for the measurement of human tears. This chapter examined the repeatability of the new instrument over multiple measurements on the same sample and over multiple days. - Chapter 5: To determine if tear film osmolality values varied significantly over the course of a normal working day in a population that was primarily free from symptoms of dry eye. - Chapter 6: To investigate the relationships between tear film osmolality and other commonly used clinical tests for dry eye disease. The clinical tests examined included various questionnaires designed to assess patient symptoms (Single Item Dry Eye Questionnaire (SIDEQ), the Ocular Surface Disease Index (OSDI), and the McMonnies Dry Eye Questionnaire (MMDEQ) and a linear analogue comfort scale (LACS)), a non-invasive tear break-up time test (NIBUT), and examination of ocular surface redness and tear ferning (TF). Secondarily to determine if the other clinical tests demonstrated significant diurnal variations over the course of a normal working day. - Chapter 7: To measure tear film osmolality in a population with mild to moderate symptoms of dry eye disease, and to compare this value with the osmolality of a population of age-matched controls without the disease. Secondarily, to investigate the relationship between tear film osmolality and patient comfort in a population with mild to moderate symptoms of dry eye disease.

Methods: - Chapter 3: Tears were collected from 40 volunteer participants with a capillary tube. Some participants were non-contact lens wearers (Non-CL), while others wore either soft or rigid contact lenses (CL). Tear film osmolality was measured with the Advanced Instruments Model 3100 Nanolitre Osmometer. - Chapter 4: Tears were collected from 10 volunteer participants using two different collection techniques. Collections were repeated on three separate days (6 study visits total); three osmolality measurements per collection were taken using the Advanced Instruments Model 3100 Nanolitre osmometer. - Chapter 5: Tears were collected from 40 volunteer participants in two separate studies (n=80 in total). Tears were collected with a capillary tube three times a day (morning, mid-day and afternoon), on two separate days (6 study visits total). Tear film osmolality was measured with the Advanced Instruments Model 3100 Nanolitre Osmometer. - Chapter 6: Clinical tests were administered and tear samples were collected using a capillary tube from 40 volunteer participants. Measurements were taken three times a day (morning, mid-day and afternoon), on two separate days (6 study visits total). Tear film osmolality was measured with the Advanced Instruments Model 3100 Nanolitre Osmometer. - Chapter 7: Participants were classified as either having dry eye disease (DE) or not having dry eye disease (NDE) based on a clinical examination that included a case history, phenol red thread test and biomicroscopy (white light and sodium fluorescein assessment). Tear samples were then collected from all participants using a capillary tube and tear film osmolality was measured with the Advanced Instruments Model 3100 Nanolitre Osmometer. Participants also completed the SIDEQ, the OSDI, and the MMDEQ.

Results: - Chapter 3: The mean tear film osmolality of the population was 298.7±11.4mOsm/Kg. CL wear (soft or rigid) did not appear to have a significant effect on tear film osmolality (CL: 298.5±11.2mOsm/Kg vs. Non-CL: 298.9±11.5mOsm/Kg), although this study was not designed to specifically look at the effects of contact lens wear on tear film osmolality. - Chapter 4: There was reasonably good concordance between measurements of tear film osmolality taken with the Advanced Instruments Model 3100 Nanolitre Osmometer (intraclass correlations range from 0.6497 (F= 0.0582) to 0.9550 (F = 0.5893)). Repeatability appeared to be affected by significant changes in ambient humidity (>10% per day). Concordance was similar with both sampling techniques. - Chapter 5: In the first study, no significant diurnal change in tear film osmolality was found (p>0.05), although a significant difference in measurements taken on Day 1 compared to Day 2 was found (p=0.040). When the first and last 10 participants enrolled were compared, the difference between days was present in the first 10 participants, but not in the last 10; it is likely that the investigator underwent a learning process during the period of the study, and that reflex tearing occurred more often in the early portion of the study compared with the latter portion. In the second study, no significant diurnal change in tear film osmolality was found (p>0.05) and no significant difference in measurements taken on Day 1 compared to Day 2 was found (p>0.05). When tear film osmolality was compared with the number of hours participants were awake, no significant correlation was found (r = 0.07044). - Chapter 6: Significant correlations were not found between tear film osmolality and SIDEQ (r = 0.1347), OSDI (r = 0.0331), MMDEQ (r = 0.2727), LACS (r = -0.1622), NIBUT (r = -0.2280), subjectively graded redness (r=-0.2280), or objectively measured redness (r = 0.1233). A weakly significant correlation was found between TF and tear film osmolality (r = 0.3978). None of the clinical measures (LACS, NIBUT, subjective or objective redness or TF) varied significantly over the course of the day. - Chapter 7: Tear film osmolality was higher in both the right (DE = 311.1±12.4mOsm/Kg, NDE = 306.2±11.2mOsm/Kg) and left eyes (DE = 313.2±11.9mOsm/Kg, NDE = 304.0±7.5mOsm/Kg) of participants, but the difference was only statistically significant in the left eye. Tear film osmolality did not correlate significantly with DE patient symptoms using any of the questionnaires (SIDEQ, OSDI, MMDEQ).

Conclusions: - Chapter 3: The Advanced Instruments Model 3100 Nanolitre Osmometer appeared to be capable of measuring tear film osmolality in a normal population. Our population mean was slightly lower than what is reported to be normal (305mOsm/Kg), but it still fell within the range of values reported as normal (297 – 318mOsm/Kg). - Chapter 4: The Advanced Instruments Model 3100 Nanolitre Osmometer demonstrated reasonably good repeatability for the measurement of human tear samples. Unfortunately, the instrumentation appeared to be affected by dramatic weather changes. Maintaining the instrument in a humidity controlled environment may resolve this problem. - Chapter 5: Tear film osmolality did not appear to vary significantly over a normal working day. Inducing reflex tearing, perhaps with an unskilled investigator collecting the tears, can be a significant source of error (as demonstrated in the first study). - Chapter 6: Tear film osmolality did not correlate well with other clinical instruments designed to assess either patient symptoms or signs of dry eye disease in a normal population. Tear film osmolality and tear ferning did demonstrate a weakly significant positive correlation. None of the clinical measures assessed demonstrated a significant diurnal variation over the course of a normal working day. - Chapter 7: Tear film osmolality appeared to be higher in participants with mild to moderate symptoms of dry eye when compared with age matched, asymptomatic controls. Tear film osmolality did not correlate well with patient symptoms in a population of mild to moderate severe dry eyed individuals.