Abstract
Purpose.:
To provide standard operating procedures (SOPs) for measuring tear inflammatory cytokine concentrations and to validate the resulting profile as a minimally invasive objective metric and biomarker of ocular surface inflammation for use in multicenter clinical trials on dry eye disease (DED).
Methods.:
Standard operating procedures were established and then validated with cytokine standards, quality controls, and masked tear samples collected from local and distant clinical sites. The concentrations of the inflammatory cytokines in tears were quantified using a high-sensitivity human cytokine multiplex kit.
Results.:
A panel of inflammatory cytokines was initially investigated, from which four key inflammatory cytokines (IL-1β, IL-6, INF-γ, and TNF-α) were chosen. Results with cytokine standards statistically satisfied the manufacturer's quality control criteria. Results with pooled tear samples were highly reproducible and reliable with tear volumes ranging from 4 to 10 μL. Incorporation of the SOPs into clinical trials was subsequently validated. Tear samples were collected at a distant clinical site, stored, and shipped to our Biomarker Laboratory, where a masked analysis of the four tear cytokines was successfully performed. Tear samples were also collected from a feasibility study on DED. Inflammatory cytokine concentrations were decreased in tears of subjects who received anti-inflammatory treatment.
Conclusions.:
Standard operating procedures for human tear cytokine assessment suitable for multicenter clinical trials were established. Tear cytokine profiling using these SOPs may provide objective metrics useful for diagnosing, classifying, and analyzing treatment efficacy in inflammatory conditions of the ocular surface, which may further elucidate the mechanisms involved in the pathogenesis of ocular surface disease.
Introduction
Cytokine profiling of body fluids has become key in the diagnosis, classification, and evaluation of inflammatory diseases.
1–3 Monitoring cytokine changes and correlating them with signs and symptoms will improve our understanding of the immune mechanisms. Ocular surface diseases are often associated with inflammation.
4–6 However, with many ocular diseases, such as dry eye disease (DED), only small volumes of nonstimulated tear samples are obtainable. To reliably quantify tear cytokines, standard operating procedures (SOPs) must be established and validated. Only then can tear cytokine profiling be incorporated into multicenter clinical trials to evaluate changes over time and across trials. Such metrics may also provide information on disease mechanisms and new approaches to treatment.
Dry eye disease is a multifactorial ocular surface disease often associated with inflammation.
7–9 The inflammatory mediators associated with DED pathogenesis can be categorized as ubiquitous inflammatory cytokines,
10,11 Th1-related cytokines,
12–14 Th17-related cytokines,
14–17 chemokines and their receptors,
14,18–22 metalloproteinase,
21,23 and secretory phospholipases.
24–26 Previous studies on DED demonstrated changes in tear cytokine expression in both humans
10,13,15,16,24 and animals,
27–31 allowing their use as a minimally invasive objective biomarker to classify severity, assess treatments, and elucidate mechanism of the disease. However, published data display significant variability (see
Supplementary Table S1).
14–16,22,32–34 Such variability could reflect biological differences among subjects as well as differences in methodology. Since variability distorts the interpretation and comparison of clinical trial outcomes, establishing SOPs will provide reliable information on tear cytokine changes.
Luminex technology (Luminex Corporation, Austin, TX) is currently one of the most frequently used cytometric bead-based methodologies for biological applications,
11,12,14,16,17,32,33,35–57 capable of simultaneously quantifying multiple targets in a single 50-μL sample with high sensitivity. To integrate tear cytokine profiles as a noninvasive objective biomarker of DED, one of the major concerns is the smaller sample volumes and potentially lower concentrations available as compared to other biological fluids, making repeat measurements unlikely, and interference between sample matrix and assay beads are common.
35,40,58 It is important to determine whether Luminex/Millipore methodologies can reliably measure tear cytokine concentrations. Also, as standard MilliPlex assay protocol requires 50 μL of sample, significant tear dilution is inevitable and the proper dilution range must be determined. The purpose of this study was to validate Luminex as a specific potential method for the analysis of tear samples from dry eye subjects.
To choose which cytokines to analyze, a literature review was performed on the roles of cytokines in DED pathogenesis and their availability for Luminex assay analysis. Previous studies indicated that IL-1β, IL-6, IFN-γ, and TNF-α have frequently displayed changes in tears of DED patients as compared to non-DED patients in same studies or research group in a year (see
Supplementary Tables S1,
S2).
14–16,22,32,33,37 Although Luminex assays can be used to quantify hundreds of molecules simultaneously from individual small-volume samples, it is not practical to assay every cytokine in one experiment.
In this study, IL-1β, IL-6, IFN-γ, and TNF-α were chosen for analysis by Luminex assays for their apparent key roles in the pathophysiology of DED and for their compatibility to fit in a single 50-μL assay. Because there is no single commercial kit available specifically tailored for tear cytokine analysis, we developed SOPs able to provide a reliable cytokine profile in multicenter randomized clinical trials (RCTs) as a biomarker of inflammation to help better understand the mechanism of the pathogenesis of ocular surface diseases and leading to novel combating strategies.
Materials and Methods
Participating Subjects
Institutional Review Board (IRB)-approved informed consents were obtained, and the tenets of the Declaration of Helsinki were followed. Tears were collected from both normal subjects and those considered to have DED per the Definition and Classification Subcommittee of the International Dry Eye WorkShop.
9 Signs and symptoms evaluated included the Ocular Surface Disease Index (OSDI) questionnaire, visual acuity assessment, Schirmer tear test, tear breakup time, corneal fluorescein staining, and conjunctival lissamine green staining, as previously described.
9,32,59
Tear Sample Collection, Storage, Shipping, and Tracking
Nonstimulated tears were collected from the lower lid margin using Drummond disposable 20-μL microcapillary tubes (Sigma-Aldrich Co. LLC, St. Louis, MO).
32,59 Precautions were taken to minimize contact between the capillary tube and the surrounding tissues. Approximately 4 to 5 μL of tears were collected from each eye in 5 minutes between 11:00 AM and 4:00 PM to minimize diurnal variation. Tears were expelled into precooled, prelabeled, individually packed and sterilized safe-lock Eppendorf tubes, sealed with Parafilm (Fisher Scientific, Inc., Pittsburgh, PA), identified with masked codes utilizing laser-printable Brady labels (Fisher Scientific, Inc.), stored on ice for <30 minutes, and then transferred to −80°C freezers until analysis. Samples collected from a distant clinical site were packed utilizing dry ice and International Air Transport Association–certified Infecon 5000 shipper kits (Fisher Scientific, Inc.), following manufacturer's instructions, and then shipped to the Mount Sinai Biomarker Laboratory on dry ice overnight. Tear samples were transferred to a −80°C freezer without thawing. All coded labels were examined for accuracy, then entered into an encrypted electronic relational database tailored for high throughput. The freezer was connected to an auto-alarm system and an emergency power backup.
Standard Tear Analysis for Cytokine Determination
Tear samples were thawed on ice and the volumes recorded. To avoid uneven evaporation, samples were kept at −80°C immediately after the volume measurements. Assay buffer was added to each sample up to 50 μL and loaded to designated wells to initiate the reaction in a 96-well plate of a High Sensitivity Human Cytokine Kit (EMD Millipore Corporation, Billerica, MA) and assayed according to the manufacturer's instructions. Plates were read on MAGPIX System (EMD Millipore Corporation); the median fluorescence intensities (MFI) were recorded, evaluated, and converted into their respective concentrations (pg/mL) using MILLIPLEX Analyst software (EMD Millipore Corporation).
Experimental Protocols
MilliPlex Assay Utilizing Commercial Standards.
Precision, Accuracy, and Sensitivity.
Multiple cytokine standard mixtures were serially diluted with assay buffer to yield final concentrations of 2000, 400, 80, 16, 3.2, 0.64, and 0.13 pg/mL, respectively. Diluted standards were assayed in four independent experiments, with duplicate measurements. Coefficient of variation (CV) was utilized to evaluate precision, calculated as standard deviation (SD) divided by the mean, and reported as a percentage (%CV). Intra-assay %CV was generated by MILLIPLEX Analyst software (EMD Millipore Corporation) from reportable results across two different concentrations of one experiment. Interassay %CV was generated from across two different concentrations (0.64 and 80 pg/mL) of four independent experiments. Accuracy was represented as the mean percentage recovery of three concentration levels of the standards ranging from 16 to 400 pg/mL in assay buffer. Sensitivity was represented as the minimum detectable concentration (MinDC, pg/mL) per manufacturer's manual. MinDC for each experiment was calculated by the MILLIPLEX Analyst software (EMD Millipore Corporation). It measures the true limits of detection by mathematically determining what the empirical MinDC would be if an infinite number of standard concentrations were run for the assay under the same conditions. MinDC of four experiments was calculated as mean + 2 SD at input of 0.00 pg/mL.
Effect of Dilution of Standards.
Variability was represented in the coefficient of multiple correlations (R 2 value). The percentage of recovery at each point was calculated as the ratio (%) of the mean of cytokine concentrations experimentally measured divided by the input.
Batch-to-Batch Repeatability.
For clinical trials, the comparability of results from a series of analyses over time must be demonstrated. To achieve this, two quality controls of known concentration ranges, as determined by the manufacturer, were measured in duplicates. QC1 represents a standard of known concentration within the lower concentration range (20–50 pg/mL), while QC2 represents the same within the higher concentration range (100–500 pg/mL). A concentration measured within the expected ranges was considered as 100% recovery.
MilliPlex Assay Utilizing Pooled Tears.
To ensure that tears can be analyzed with the same precision as commercial samples, the following experiments were performed.
Precision and Repeatability.
Two pooled tear samples with large volumes (designated A and B) were independently collected from normal subjects and assayed. Ten microliters of tears from each was diluted into 40 μL assay buffer to make a 5× dilution. The dilutions were measured seven times (A) and eight times (B) as described above, and the mean ± SD of the concentrations was presented.
Effect of Tear Dilution.
To evaluate the repeatability of tear dilutions, the two pooled tear samples (A and B) were each analyzed with 2, 4, 6, 8, or 10 μL tears diluted to 50 μL with assay buffer. This experiment had seven (A) or eight (B) replicates. The concentrations of each dilution are presented as the mean ± SD.
Effect of Tears on Assay Matrix: Spike Recovery Assay of Low, Middle, and High Levels of Standards.
To determine if tears interfere with Luminex assay results, a spike recovery experiment was conducted using groups of 50 μL/well samples, each containing 5 μL of one of the two pooled tear samples (designated C and D) and 45 μL of assay buffer with cytokine standards at various concentrations (0, 8, 40, or 200 pg/mL, respectively). Samples (six replicates) with cytokine standards alone were included as controls. The percentage of recovery of the standards in each group was calculated as the ratio of the mean of observed concentration over the expected concentration of each analyte.
Availability of Needed Tear Volume per Subject.
To determine the ability to obtain sufficient tear volume from clinical subjects (DED or non-DED) using the SOPs, tear samples from right (OD) and left (OS) eye per subject were either collected separately (OD or OS tears) or pooled together (OD + OS tears) and their volumes determined.
Batch-to-Batch Repeatability.
To create tear standards consistent over time and across batches, two pooled tear samples, designated QC3A and QC3B, were collected and stocks of 10-μL aliquots produced. Aliquots of each were measured twice, and the percent recovery and %CV were calculated, as described for the QC1 and QC2 standards, to yield the actual concentrations for batch-to-batch comparison. The unused aliquots, sufficient for 2 years of use, were stored in a −80°C freezer. Care was taken to prevent tear standards from running out before completion of a trial. Should the trial be completed for more than 2 years, new aliquots of QC3A and QC3B must be made the same way described above before the current aliquots are used up or expired.
Ability to Collect and Analyze Tear Samples From a Distant Site.
To determine the ability to teach a distant site to obtain, store, and ship tear samples to the Biomarker Laboratory and demonstrate the ability of the laboratory to receive, track, and analyze received samples for multicenter clinical trials, interaction with an out-of-state site was instituted. Teaching was accomplished by a Web-based presentation followed by conference calls to review SOP details and answer questions. A total of 40 tear samples, 20 from contact lens wearers and 20 from age-matched noncontact lens wearers, were collected under an IRB-approved study and shipped to and assayed at Mount Sinai according to our SOPs.
Ability to Incorporate Tear Collection and Cytokine Analysis Into a Double-Blind RCT.
Tears pooled from the right (OD) and left (OS) eyes of each DED subject in an IRB- and FDA-approved double-blind RCT study of systemic omega (ω)3 treatment were stored and analyzed at baseline and after 3 months of treatment according to the SOPs described above.
Statistical Analysis
Student's t-test (parametric), Wilcoxon-Mann-Whitney rank-sum test (nonparametric), and one-way analysis of variance (one-way ANOVA) on SPSS (version 17; SPSS, Inc., Chicago, IL) were run. Parametric results were presented as mean ± SD. Two-tailed significance was established at a confidence level of 0.05 > P > 0.95.
Results
MilliPlex Assay Utilizing Commercial Standards
Independent experiments were run with serial dilutions of standards and known concentration controls, each with duplicate measurements. Typical standard curves are shown in
Figure 1. The intra-%CV is less than 1.7% and
R 2 is approximately 1, making the “data fit” almost linear. Although the slopes between the initial baseline and an arbitrary linearity line matching the early to middle logarithm phase part of the curves were nearly the same for all four measured cytokines, the curve of IFN-γ displayed a bigger lag phase, which increased the maximum detecting capacity over 2000 pg/mL.
Precision, Accuracy, and Sensitivity.
As shown in
Table 1, a broad detectable range was achieved for each cytokine. The measured concentrations were almost the same as those of the standards (
R 2 ≥ 0.99). The inter-%CV of the four experiments was 15.86 for IL-1β, 20.12 for IL-6, 40.04 for IFN-γ, and 39.47 for TNF-α. The MinDCs of the four experiments were 0.13 for IL-1β, 1.22 for IL-6, 1.21 for IFN-γ, and 0.10 (pg/mL) for TNF-α.
Table 1 Summary of Precision and Accuracy Determination With Four Cytokine Standards
Table 1 Summary of Precision and Accuracy Determination With Four Cytokine Standards
| | Experiment 1 | Experiment 2 | Experiment 3 | Experiment 4 | Interassay, %CV | MinDC, pg/mL |
| Detectable Ranges, pg/mL | %CV | R 2 | Detectable Ranges, pg/mL | %CV | R 2 | Detectable Ranges pg/mL | %CV | R 2 | Detectable Ranges, pg/mL | %CV | R 2 |
| IL-1β | 0.11–425.52 | 4.15 | 0.99 | 0.06–2136.32 | 0.74 | 1.00 | 0.07–9,575.60 | 1.48 | 1.00 | 0.02–1523.07 | 0.71 | 1.00 | 15.86 | 0.13 |
| IL-6 | 0.10–258.28 | 5.61 | 0.99 | 0.11–2302.48 | 1.17 | 1.00 | 0.79–10,967.81 | 1.45 | 1.00 | 0.03–1587.28 | 1.66 | 1.00 | 20.12 | 1.22 |
| IFN-γ | 0.14–1574.98 | 1.96 | 0.99 | 0.07–1991.70 | 0.95 | 1.00 | 0.40–11,101.25 | 0.95 | 1.00 | 0.02–2230.03 | 0.95 | 1.00 | 40.04 | 1.21 |
| TNF-α | 0.02–213.42 | 0.52 | 1.00 | 0.05–3996.88 | 0.32 | 1.00 | 0.08–7,547.66 | 2.16 | 0.99 | 0.04–1681.72 | 1.03 | 1.00 | 39.47 | 0.10 |
Table 2 Recovery Assays of Standards
Table 2 Recovery Assays of Standards
| Measured | IL-1β | IL-6 | IFN-γ | TNF-α |
| Input | Mean ± SD, pg/mL | N* | % Recovery | Mean ± SD, pg/mL | N | % Recovery | Mean ± SD, pg/mL | N | % Recovery | Mean ± SD, pg/mL | N | % Recovery |
| 0 | 0.03 ± 0.05 | 8 | NA | 0.24 ± 0.44 | 8 | NA | 0.25 ± 0.48 | 8 | NA | 0.02 ± 0.04 | 8 | NA |
| 0.13 | 0.08 ± 0.07 | 8 | 62 | 0.13 ± 0.17 | 8 | 97 | 0.23 ± 0.31 | 8 | 175 | 0.09 ± 0.08 | 8 | 72 |
| 0.64 | 0.66 ± 0.16 | 8 | 103 | 0.65 ± 0.15 | 8 | 102 | 0.6 ± 0.27 | 8 | 94 | 0.74 ± 0.13 | 8 | 115 |
| 3.2 | 3.39 ± 0.5 | 8 | 106 | 3.91 ± 0.52 | 8 | 122 | 3.08 ± 1.67 | 8 | 96 | 3.46 ± 0.53 | 8 | 108 |
| 16 | 15.89 ± 3.67 | 8 | 99 | 14.17 ± 5.29 | 8 | 89 | 16.09 ± 5.66 | 8 | 101 | 15.72 ± 5.72 | 8 | 98 |
| 80 | 72.04 ± 12.28 | 8 | 90 | 79.16 ± 21.24 | 8 | 99 | 83.47 ± 21.52 | 8 | 104 | 97.92 ± 62.33 | 8 | 122 |
| 400 | 437.23 ± 66.7 | 8 | 109 | 416.3 ± 182.4 | 8 | 104 | 401.99 ± 81.39 | 8 | 100 | 400.49 ± 164.95 | 8 | 100 |
| 2000 | 2086.21 ± 674.69 | 8 | 104 | 2149.92 ± 526.48 | 8 | 107 | 1836.8 ± 222.8 | 7* | 92 | 1783.36 ± 1209.75 | 8 | 89 |
Effect of Dilution of Standards on Accuracy.
The mean of cytokine concentrations, the standard deviation of the mean, and the percentage recoveries are shown in
Table 2. The recoveries at each diluting level over eight replicates were in the 80% to 120% range, although a slightly larger variance was seen at the lowest point (0.13 pg/mL), attesting to the approach toward the minimal detectable limit. The linearity of measured concentrations of all cytokines across dilutions was a near-perfect fit to the expected values of the standards (
R 2 > 0.99).
Batch-to-Batch Repeatability.
The measured mean concentrations of IL-6 and INF-γ, both QC1 and QC2, were within the expected ranges (see
Table 3), indicating 100% recovery. The mean recovery values of TNF-γ and IL-1β in QC1 were also within the expected ranges, while those in QC2 were slightly above the upper ranges (105% for TNF-α and 129% for IL-1β). The %CVs were no more than 11%. These results met the quality control criteria set by the manufacturer.
Table 3 A Representative Experiment With Three Quality Controls of Four Cytokines at Known Concentration Ranges
Table 3 A Representative Experiment With Three Quality Controls of Four Cytokines at Known Concentration Ranges
| ID | Volume, μL | IL-1β | IL-6 | INF-γ | TNF-α |
| Expected, pg/mL | Recovery, % | % CV | Expected, pg/mL | Recovery, % | % CV | Expected, pg/mL | Recovery, % | % CV | Expected, pg/mL | Recovery, % | % CV |
| QC1 | 50 | 20–42 | 100 | 0.43 | 22–46 | 100 | 7.41 | 24–50 | 100 | 10.82 | 21–44 | 100 | 5.23 |
| QC2 | 50 | 154–320 | 129 | 0.87 | 177–367 | 100 | 3.06 | 179–372 | 100 | 2.21 | 168–348 | 105 | 3.2 |
| QC3A | 50 | 0.07–0.14 | 100 | 19.4 | 0.42–0.70 | 103 | 17.6 | 3.75–5.23 | 100 | 19.12 | 0.51–0.65 | 100 | 17.87 |
| QC3B | 50 | 0.10–0.42 | 117 | 21.66 | 0.60–1.04 | 107 | 23.59 | 4.82–6.96 | 100 | 19.92 | 0.64–0.88 | 108 | 17.92 |
MilliPlex Assay Utilizing Pooled Tears
Precision and Repeatability.
With the pooled tear sample A, the %CVs of the four assayed cytokines were shown to be less than 20% at most of the dilution points. The mean %CVs of all the 36 measures across dilutions were also less than 20% (see
Table 4). The concentrations were 0.11 ± 0.03 for IL-1β, 0.56 ± 0.14 for IL-6, 4.49 ± 0.74 for IFN-γ, and 0.58 ± 0.07 pg/mL TNF-α. When 4 to 10 μL tears was measured, the
P values were larger than 0.05 for all cytokines, indicating no statistical difference. Similar results were seen with pooled tear sample B.
Table 4 Levels of Four Cytokines in Two Pooled Normal Tear Samples
Table 4 Levels of Four Cytokines in Two Pooled Normal Tear Samples
| Sample ID | Tear Vol., μL | Dilution Factor | Number of Repeats | IL-1β | IL-6 | IFN-γ | TNF-α |
| Mean ± SD | % CV | P Value | Mean ± SD | % CV | P Value | Mean ± SD | % CV | P Value | Mean ± SD | % CV | P Value |
| A1 | 2 | 25 | 7 | <0.02 ± 0.01 | 18.16 | <0.05* | 0.36 ± 0.11 | 17.7 | >0.05 | 3.51 ± 1.34 | 22.54 | >0.05 | 0.46 ± 0.1 | 14.75 | >0.05 |
| A2 | 4 | 12.5 | 8 | 0.07 ± 0.08 | 23.79 | >0.05 | 0.53 ± 0.23 | 28.18 | >0.05 | 4.8 ± 1.88 | 27.9 | >0.05 | 0.59 ± 0.19 | 25.32 | >0.05 |
| A3 | 6 | 8.3 | 7 | 0.1 ± 0.06 | 19.89 | >0.05 | 0.53 ± 0.11 | 13.38 | >0.05 | 4.32 ± 1.59 | 24.9 | >0.05 | 0.59 ± 0.12 | 15.9 | >0.05 |
| A4 | 8 | 6.3 | 7 | 0.12 ± 0.09 | 19.06 | >0.05 | 0.68 ± 0.12 | 12.99 | >0.05 | 5.53 ± 0.71 | 9.66 | >0.05 | 0.65 ± 0.12 | 14.5 | >0.05 |
| A5 | 10 | 5 | 7 | 0.14 ± 0.08 | 16.1 | >0.05 | 0.72 ± 0.16 | 15.73 | >0.05 | 4.31 ± 0.69 | 10.61 | >0.05 | 0.63 ± 0.15 | 18.87 | >0.05 |
| SUM | | | 36 | 0.11 ± 0.03 | 19.4 | | 0.56 ± 0.14 | 17.6 | | 4.49 ± 0.74 | 19.12 | | 0.58 ± 0.07 | 17.87 | |
| B1 | 2 | 25 | 8 | 0.03 ± 0.04 | 16.93 | <0.05† | 0.5 ± 0.16 | 20.64 | <0.05† | 4.11 ± 1.1 | 17 | <0.05* | 0.61 ± 0.15 | 19.48 | >0.05 |
| B2 | 4 | 12.5 | 8 | 0.25 ± 0.11 | 16.82 | >0.05 | 0.79 ± 0.2 | 19 | >0.05 | 5.76 ± 1.16 | 15.11 | >0.05 | 0.73 ± 0.11 | 12.39 | >0.05 |
| B3 | 6 | 8.3 | 8 | 0.29 ± 0.18 | 25.65 | >0.05 | 0.86 ± 0.25 | 21.65 | >0.05 | 6.24 ± 1.62 | 20.85 | >0.05 | 0.77 ± 0.17 | 18.16 | >0.05 |
| B4 | 8 | 6.3 | 8 | 0.24 ± 0.1 | 16.22 | >0.05 | 0.82 ± 0.21 | 19.08 | >0.05 | 6.51 ± 1.34 | 16.99 | >0.05 | 0.76 ± 0.1 | 11.01 | >0.05 |
| B5 | 10 | 5 | 8 | 0.49 ± 0.31 | 32.7 | >0.05 | 1.11 ± 0.49 | 37.56 | >0.05 | 6.82 ± 2.34 | 29.67 | >0.05 | 0.95 ± 0.3 | 28.57 | >0.05 |
| SUM | | | 40 | 0.26 ± 0.16 | 21.66 | | 0.82 ± 0.22 | 23.59 | | 5.89 ± 1.07 | 19.92 | | 0.76 ± 0.12 | 17.92 | |
Effect of Tear Dilution.
A preliminary experiment using 2.5 to 10 μL of a pooled tear sample generated similar results (P > 0.05). This range is similar to the tear volumes usually obtained from DED patients. We therefore focused on two additional pooled tear groups (A and B) with more repeats (eight [A], seven [B]) to further narrow down the proper dilution ranges (5- to 25-fold). The mean concentrations displayed no statistically significant difference across the dilutions with tear volumes from 4 to 10 μL (P > 0.05), but showed significant differences from those with 2 μL (dilution factor of 25).
Upon normalization, the mean concentrations of all cytokines were plotted against tear volumes semilogarithmically, resulting in a nearly straight line with ≈0 slope and reasonable SD values (except for 2 μL points; see
Fig. 2). Thus, the minimal tear volume required for accuracy is greater than 2 μL and, per our protocol, 4 μL would be optimal.
Availability of Needed Tear Volume per Subject.
To test the ability of collecting sufficient tears for analysis, a total of 193 tear samples were measured (121 samples from DED subjects, 72 from non-DED; see
Table 5).
| | Mean, μL | SD | N | % With Tear Volume ≥ 4 μL |
| DED subjects |
| OD or OS tears | 2.64 | 1.11 | 29 | 14 |
| OD + OS tears | 14.18 | 11.95 | 92 | 81 |
| Subtotal | | | 121 | |
| Non-DED subjects |
| OD or OS tears | 4 | 2.19 | 44 | 50 |
| OD + OS tears | 11.51 | 6.43 | 28 | 93 |
| Subtotal | | | 72 | |
Among the DED tears, 29 samples were separately collected and measured from either right (OD) or left (OS) eyes (“OD or OS tears”). The mean tear volume ± SD was 2.64 ± 1.11 μL (range, 0–5 μL); only 14% of the samples had tear volumes ≥ 4 μL. However, when OD and OS tears were pooled together (“OD + OS tears”), the mean ± SD went up to 14.18 ± 11.95 μL (n = 92; range, 0–50 μL); over 81% samples had tear volumes ≥ 4 μL.
Among the non-DED tears, both the nonpooled and pooled OD + OS tears had volumes ≥ 4.0 μL (see
Table 5).
A pooled OD + OS tear sample, from either DED or non-DED subjects, would therefore likely (≥81%) be sufficient for analysis utilizing the MILLIPLEX MAP High Sensitivity Human Cytokine Magnetic Bead Panel (EMD Millipore Corporation). Since higher concentrations are expected in patients with inflammatory diseases, even smaller tear volumes might be accurately measured.
Batch-to-Batch Repeatability.
The mean percent recovery of QC3A or QC3B was within the 80% to 120% range (see
Table 3), and the mean %CVs were ≈20%.
Effect of Tears on Assay Matrix: Spike Recovery Assay of Low, Middle, and High Levels of Standards in Tears.
In “Standards alone” groups (S0–S3) of the spike recovery results (
Table 6), an overall 99% recovery was achieved (range, 84%–121%), similar to published results.
35,40 In “Tears alone” groups (S0-C, S0-D), batches of pooled tears were similar in all cytokines, while in “Standards + Tear-C” or “Standards + Tear-D” groups (S1-C, S2-C, S3-C, S1-D, S2-D, S3-D), the overall recovery was ≈40% (range, 15%–71%), with a trend toward decreasing recovery as the standards' concentrations increased.
Table 6 Spike Recovery With Tears and Cytokine Standards
Table 6 Spike Recovery With Tears and Cytokine Standards
| | Assays | IFN-γ | IL-1β | IL-6 | TNF-α |
| Group ID | Tears, μL | Spiked Standards, pg/mL | Mean, pg/mL | SD | % Recovery | Mean, pg/mL | SD | % Recovery | Mean, pg/mL | SD | % Recovery | Mean, pg/mL | SD | % Recovery |
| Standards alone, n = 6 | S0 | 0 | 0 | <Threshold | NA | NA | <Threshold | NA | NA | <Threshold | NA | NA | <Threshold | NA | NA |
| S1 | 0 | 8 | 8.21 | 5.8 | 103 | 8.24 | 2.42 | 103 | 7.97 | 3.49 | 100 | 7.64 | 3.11 | 95 |
| S2 | 0 | 40 | 48.58 | 16.97 | 121 | 39.68 | 11.28 | 99 | 34.91 | 15.61 | 87 | 33.51 | 15.48 | 84 |
| S3 | 0 | 200 | 196.48 | 37.91 | 98 | 233.81 | 38.59 | 117 | 171.86 | 46.02 | 86 | 240.9 | 130.44 | 120 |
| Standard + tear-C, n = 6 | S0-C | 5 | 0 | 18.63 | 5.18 | NA | 0.64 | 0.28 | NA | 3.87 | 1.44 | NA | 1.54 | 0.33 | NA |
| S1-C | 5 | 8 | 18.84 | 3.78 | 71 | 4.69 | 0.62 | 54 | 6.84 | 1.08 | 58 | 2.47 | 0.17 | 26 |
| S2-C | 5 | 40 | 26.46 | 8.11 | 45 | 18.4 | 3.76 | 41 | 20.01 | 7.39 | 43 | 6.39 | 0.41 | 15 |
| S3-C | 5 | 200 | 95.94 | 17.86 | 42 | 94.68 | 19.06 | 43 | 83.61 | 22.91 | 38 | 38.41 | 0.34 | 19 |
| Standard + tear-D, n = 6 | S0-D | 5 | 0 | 12.83 | 5.31 | NA | 0.41 | 0.14 | NA | 2.89 | 0.62 | NA | 1.24 | 0.25 | NA |
| S1-D | 5 | 8 | 14.72 | 5.16 | 71 | 5.25 | 1.5 | 63 | 6.73 | 2.23 | 62 | 2.3 | 0.25 | 25 |
| S2-D | 5 | 40 | 28.51 | 11.03 | 52 | 18.02 | 3.61 | 40 | 18.5 | 5.91 | 40 | 6.2 | 0.34 | 15 |
| S3-D | 5 | 200 | 101.22 | 20.38 | 44 | 89.97 | 21.15 | 41 | 92.58 | 34.14 | 42 | 40.9 | 0.4 | 20 |
Ability to Collect and Analyze Tear Samples From a Distant Site
There was no significant difference in tear volumes between contact lens wearers and nonwearers, or between the volumes reported by the collectors and those measured in the lab; there was no evaporation loss during tear transfer and storage.
All samples were received in good condition and sufficient volumes for analysis. The concentrations of inflammatory cytokines IL-6, IFN-γ, and TNF-α were not significantly different (
P > 0.05) between the groups, and those of IL-1β were only slightly higher in the contact lens wearer group (0.05 >
P > 0.01; see
Table 7).
Table 7 Tear Cytokine Concentrations of Samples From Distant Site (pg/mL)
Table 7 Tear Cytokine Concentrations of Samples From Distant Site (pg/mL)
| Cytokine | Normal | Contact Lens Wearers | P Value |
| Mean ± SD | N | Mean ± SD | N |
| IL-1β | 7.42 ± 5.62 | 20 | 15.56 ± 12.94 | 20 | 0.016* |
| IL-6 | 13.43 ± 8.74 | 20 | 14.86 ± 0.19 | 20 | 0.474 |
| IFN-γ | 0.27 ± 0.88 | 20 | 1.25 ± 3.69 | 20 | 0.261 |
| TNF-α | 7.46 ± 8.74 | 20 | 14.82 ± 15.8 | 20 | 0.078 |
Ability to Incorporate Tear Collection and Cytokine Analysis Into a Double-Blind RCT
To demonstrate the ability to incorporate tear sampling and cytokine analysis into a double-blind RCT, a feasibility double-blind RCT of ω3 for treatment of DED was performed utilizing the SOPs. There were no significant tear volume changes between the baseline and 3-month collections.
Tear samples were analyzed for cytokine concentrations in a masked fashion. Cytokine results were analyzed based on treatment after all patients completed the protocol and their data were entered into the relational database. The levels in three out of four selected tear cytokines displayed a tendency to decrease upon ω3 treatment: IL-1β (4×↓), IFN-γ (3×↓), TNF-α (4×↓). Interleukin-6 was increased (4×↑) as compared to values in placebo subjects (
Fig. 3; for details see
Supplementary Table S3). These preliminary results suggest that ω3 may decrease inflammatory cytokines or delay progression of DED inflammation as compared to that in placebo controls; likely larger subject populations and longer treatment would be needed to determine efficacy of ω3 on reducing tear cytokine concentrations.
Discussion
Clinical trials of DED often provide less than clear determination of efficacy, likely due to outcomes blurred by sources of bias in outcome measures, such as observer rating of corneal staining. Inclusion of minimally invasive objective metrics such as tear inflammatory cytokine profiles will aid in developing and determining the efficiency of new treatments as well as understanding changes in the ocular surface for specific external diseases, such as DED. The research reported here validates the SOPs needed to include tear sampling and reliable cytokine analysis in multicenter clinical trials.
To reliably include tear profiles as an ocular surface inflammatory biomarker, we first determined the sensitivities and reliabilities of a MILLIPLEX MAP High Sensitivity Human Cytokine Magnetic Bead Panel (EMD Millipore Corporation) for tear cytokine analysis. We evaluated the precision, repeatability, accuracy, sensitivity, effect of dilution, and batch-to-batch repeatability of the MILLIPLEX MAP High Sensitivity Human Cytokine Magnetic Bead Panel (EMD Millipore Corporation) using the commercial standards, then confirmed the ability of the kit to process cytokine analysis of tears using the SOPs with pooled tears. Given that DED is likely to have multifactoral pathogenesis, simultaneous analysis of several key cytokines is essential. Given that tear samples, especially those obtained from individuals with DED, are likely small in volume and often nonrepeatable in assay, a highly sensitive methodology with reliable results is essential. Although methods such as ELISA
60–62 and proteomic arrays
60,63 have been used for multiple cytokine quantification, both require relatively larger sample volumes and are less sensitive than Luminex. Some recently introduced super high-sensitivity–high-throughput technologies, such as nanofibril-based nanoarray,
64,65 are still under development. Our results demonstrated that the MILLIPLEX MAP High Sensitivity Human Cytokine Magnetic Bead Panel (EMD Millipore Corporation) meets all the above requirements and is appropriate for the SOPs.
We then demonstrated that the profile of a panel of inflammatory cytokines (IL-1β, IL-6, IFN-γ, and TNF-α) can be reliably used as a biomarker of DED. Extensive review of literature revealed that these four cytokines are the most frequently detected as increased in tears of DED subjects (
Supplementary Tables S1,
S2),
14–16,22,32,33,37 play key roles in DED pathogenesis, correlate with mechanisms of immune modulation (such as IFN-γ increases in Th1 and IL-6 increases in Th17), and can be fit into one assay for simultaneous quantification with the highest sensitivity. Although the SOPs were optimized with only four cytokines, the chosen panel could be expanded and tailored to fit any specific patient group or disease type.
Tear matrices are known to contain various substances that may interfere with Luminex assays. The absolute concentration of these substances and how they influence the accuracy of tear cytokine measurements may vary. Our spike recovery results in
Table 6 indicate that, with just 5 μL of tears, recovery results were significantly decreased. Similar interference was also observed by other peers using Luminex assay with tear or serum samples, or even with cytokine standards, probably due to multiple factors such as cytokine structures, and/or those of beads and sample matrix that could change the microreaction environment.
35,40,58,66 Our results imply that the interfering effects vary: The most were seen with TNF-α and the least with IFN-γ. The interfering effects are not overcome by simply increasing concentrations of the standards, suggesting that factors other than inhibitors may be also involved, which is likely common in Luminex beads assay with many biological fluids.
35,40,58,66 Therefore, a “specialized” buffering system designed specifically to work for a particular sample type such as a tear sample, and a “prefiltering” step prior to Luminex assay to remove or reduce the interfering substances, would be helpful.
Furthermore, data from published reports that include both DED and non-DED results indicate that the mean concentrations of these four inflammatory cytokines in normal individuals are 39.0 ± 23.6 pg/mL for IL-1β, 42.2 ± 23.6 pg/mL for IL-6, 24.0 ± 18.0 pg/mL for IFN-γ, and 58.3 ± 36.9 pg/mL for TNF-α (see
Supplementary Table S2).
14–16,22,32,33,37 Results shown in
Figure 2 and
Table 4 were highly reproducible with volumes equivalent to 4 to 10 μL of normal stimulated tears using the MILLIPLEX MAP High Sensitivity Human Cytokine Magnetic Bead Panel (EMD Millipore Corporation). Taking these data together, we estimate that 4 μL of normal tears could provide reliable results for all for cytokines. To support this, volumes of 193 nonstimulated tear samples from DED and non-DED subjects were measured. Results indicated that the mean tear volume of DED subjects was below 4 μL (2.64 ± 1.11 μL,
n = 29) when tears of the left or right eye per subject were collected and measured separately. However, the mean tear volume would go up over 10 μL when a pooled tear sample of both eyes per subject was used (14.18 ± 11.95 μL,
n = 92). Clinically, at least 4 μL pooled tears could be collected from 93% of normal individuals and 81% of DED subjects by the SOPs (see
Table 5). Therefore, a pooled tear sample from an individual is likely to be sufficient for MilliPlex analysis. It needs to be mentioned that the benchmark of 4 μL was justified based on measurements of the pooled normal stimulated tears, in which the concentrations of the cytokines could be even lower than those reported in nonstimulated tears of non-DED controls (
Supplementary Tables S1,
S2).
14–16,22,32,33,37 Since tear samples from DED subjects are expected to have higher concentrations of inflammatory cytokines than those from normals, an even higher percentage of samples from DED subjects could likely be reliably evaluated.
Finally, we demonstrated that the SOPs for tear cytokine analysis could be reasonably incorporated into multicenter, double-blind RCTs. To achieve this goal, all involved sites would need to have trained personnel who could reliably collect, label, store, and ship samples to the Biomarker Laboratory for processing. The latter would then have to demonstrate the ability to receive, track, and analyze the masked samples and store the results electronically to send to a coordinate center for analysis. We demonstrated that, using the SOPs, a distant clinic site with Web-based education could obtain good-quality tear samples for inflammatory cytokine quantification. We also successfully incorporated the SOPs into a double-blind RCT of DED and traced the trend in cytokine changes between baseline and after 3 months of ω3 treatment.
We also incorporated into our SOPs the previously established information important for tear cytokine profiling: collection of tears at the same time of the day for each subject given our knowledge of diurnal rhythm variation in tear cytokine concentration
36 ; use of capillary tubes rather than Schirmer strips for tear collection; maximum time (30 minutes) to keep tears on ice after collection and before freezing to ensure stability of cytokines; and avoidance of long-term tear sample storage at −80°C (less than 2 years) and multiple frozen–thaw cycles (no more than 2 cycles).
35 Critical steps in the SOPs to ensure reliable results are detailed in
Figure 4. The SOPs ensure that results are reliable and help to eliminate sources of variation and bias found in other outcome measures typically used in DED clinical trials.
In summary, we have established and validated SOPs for human tear inflammatory cytokine assessment for use in multicenter clinical trials and demonstrated that the profile of tear inflammatory cytokines IL-1β, IL-6, IFN-γ, and TNF-α can be used as a biomarker to monitor inflammatory diseases of the ocular surface such as DED. Tear cytokine biomarkers may provide critical information to fill the gap between clinical treatment modalities and the mechanism(s) involved in DED pathogenesis and response to treatment.
Supplementary Materials
Acknowledgments
The authors thank Brittlyn Pearlman, Peter Dentone, Geoffrey Raynor, and Karen Fernandez at the Icahn School of Medicine at Mount Sinai for their assistance and the efforts of Padmabriya Ranamoorthy, Jason Nichols, Jillian Meadows, and Kelly Nichols at The Ohio State University in providing tear samples from a distant clinical site.
Supported by National Institutes of Health Fund R34EY017626 and the Martin and Toni Sosnoff Foundation. The authors alone are responsible for the content and writing of the paper.
Disclosure: Y. Wei, None; N. Gadaria-Rathod, None; S. Epstein, None; P. Asbell, None
References
Bjorkbacka
H
Fredrikson
GN
Nilsson
J.
Emerging biomarkers and intervention targets for immune-modulation of atherosclerosis - a review of the experimental evidence.
Atherosclerosis
.
2013;
227:
9–17.
[CrossRef] [PubMed]
Desai
D
Brightling
C.
Cytokines and cytokine-specific therapy in asthma.
Adv Clin Chem
.
2012;
57:
57–97.
[PubMed]
Babic
GS
Zlatanovic
G
Jocic
JD
Cekic
S
Vujanovic
M.
Therapeutical approach to dry eye syndrome [in Serbian].
Med Pregl
.
2010;
63:
793–800.
[CrossRef] [PubMed]
Chen
Y
Chauhan
SK
Soo Lee
H
Saban
DR
Dana
R.
Chronic dry eye disease is principally mediated by effector memory Th17 cells
[published online ahead of print April 10, 2013].
Mucosal Immunol
.
doi:
10.1038/mi.2013.20
.
Stern
ME
Schaumburg
CS
Pflugfelder
SC.
Dry eye as a mucosal autoimmune disease.
Int Rev Immunol
.
2013;
32:
19–41.
[CrossRef] [PubMed]
Ueta
M
Kinoshita
S.
Ocular surface inflammation is regulated by innate immunity.
Prog Retin Eye Res
.
2012;
31:
551–575.
[CrossRef] [PubMed]
Stevenson
W
Chauhan
SK
Dana
R.
Dry eye disease: an immune-mediated ocular surface disorder.
Arch Ophthalmol
.
2012;
130:
90–100.
[CrossRef] [PubMed]
Pflugfelder
SC
Stern
ME.
Immunoregulation on the ocular surface: 2nd Cullen Symposium.
Ocul Surf
.
2009;
7:
67–77.
[CrossRef] [PubMed]
The definition and classification of dry eye disease: report of the Definition and Classification Subcommittee of the International Dry Eye WorkShop (2007).
Ocul Surf
.
2007;
5:
75–92.
[CrossRef] [PubMed]
Boehm
N
Riechardt
AI
Wiegand
M
Pfeiffer
N
Grus
FH.
Proinflammatory cytokine profiling of tears from dry eye patients by means of antibody microarrays.
Invest Ophthalmol Vis Sci
.
2011;
52:
7725–7730.
[CrossRef] [PubMed]
Narayanan
S
Corrales
RM
Farley
W
McDermott
AM
Pflugfelder
SC.
Interleukin-1 receptor-1-deficient mice show attenuated production of ocular surface inflammatory cytokines in experimental dry eye.
Cornea
.
2008;
27:
811–817.
[CrossRef] [PubMed]
Corrales
RM
Villarreal
A
Farley
W
Stern
ME
Li
DQ
Pflugfelder
SC.
Strain-related cytokine profiles on the murine ocular surface in response to desiccating stress.
Cornea
.
2007;
26:
579–584.
[PubMed]
Paiva
CS
Pflugfelder
SC.
Rationale for anti-inflammatory therapy in dry eye syndrome.
Arq Bras Oftalmol
.
2008;
71:
89–95.
[CrossRef] [PubMed]
Lam
H
Bleiden
L
de Paiva
CS
Farley
W
Stern
ME
Pflugfelder
SC.
Tear cytokine profiles in dysfunctional tear syndrome.
Am J Ophthalmol
.
2009;
147:
198–205,
e191.
[CrossRef] [PubMed]
Yoon
KC
Jeong
IY
Park
YG
Yang
SY.
Interleukin-6 and tumor necrosis factor-alpha levels in tears of patients with dry eye syndrome.
Cornea
.
2007;
26:
431–437.
[CrossRef] [PubMed]
Enriquez-de-Salamanca
A
Castellanos
E
Stern
ME
Tear cytokine and chemokine analysis and clinical correlations in evaporative-type dry eye disease.
Mol Vis
.
2010;
16:
862–873.
[PubMed]
De Paiva
CS
Hwang
CS
Pitcher
JD III
Age-related T-cell cytokine profile parallels corneal disease severity in Sjogren's syndrome-like keratoconjunctivitis sicca in CD25KO mice.
Rheumatology (Oxford)
.
2010;
49:
246–258.
[CrossRef] [PubMed]
Yoon
KC
Park
CS
You
IC
Expression of CXCL9, −10, −11, and CXCR3 in the tear film and ocular surface of patients with dry eye syndrome.
Invest Ophthalmol Vis Sci
.
2010;
51:
643–650.
[CrossRef] [PubMed]
Gulati
A
Sacchetti
M
Bonini
S
Dana
R.
Chemokine receptor CCR5 expression in conjunctival epithelium of patients with dry eye syndrome.
Arch Ophthalmol
.
2006;
124:
710–716.
[CrossRef] [PubMed]
El Annan
J
Chauhan
SK
Ecoiffier
T
Zhang
Q
Saban
DR
Dana
R.
Characterization of effector T cells in dry eye disease.
Invest Ophthalmol Vis Sci
.
2009;
50:
3802–3807.
[CrossRef] [PubMed]
Corrales
RM
Stern
ME
De Paiva
CS
Welch
J
Li
DQ
Pflugfelder
SC.
Desiccating stress stimulates expression of matrix metalloproteinases by the corneal epithelium.
Invest Ophthalmol Vis Sci
.
2006;
47:
3293–3302.
[CrossRef] [PubMed]
Acera
A
Rocha
G
Vecino
E
Lema
I
Duran
JA.
Inflammatory markers in the tears of patients with ocular surface disease.
Ophthalmic Res
.
2008;
40:
315–321.
[CrossRef] [PubMed]
Chotikavanich
S
de Paiva
CS
de Q
Li
Production and activity of matrix metalloproteinase-9 on the ocular surface increase in dysfunctional tear syndrome.
Invest Ophthalmol Vis Sci
.
2009;
50:
3203–3209.
[CrossRef] [PubMed]
Chen
D
Wei
Y
Li
X
Epstein
S
Wolosin
JM
Asbell
P.
sPLA2-IIa is an inflammatory mediator when the ocular surface is compromised.
Exp Eye Res
.
2009;
88:
880–888.
[CrossRef] [PubMed]
Wei
Y
Epstein
SP
Fukuoka
S
Birmingham
NP
Li
XM
Asbell
PA.
sPLA2-IIa amplifies ocular surface inflammation in the experimental dry eye (DE) BALB/c mouse model.
Invest Ophthalmol Vis Sci
.
2011;
52:
4780–4788.
[CrossRef] [PubMed]
Wei
Y
Pinhas
A
Liu
Y
Epstein
S
Wang
J
Asbell
P.
Isoforms of secretory group two phospholipase A (sPLA2) in mouse ocular surface epithelia and lacrimal glands.
Invest Ophthalmol Vis Sci
.
2012;
53:
2845–2855.
[CrossRef] [PubMed]
Nguyen
CQ
Yin
H
Lee
BH
Chiorini
JA
Peck
AB.
IL17: potential therapeutic target in Sjogren's syndrome using adenovirus-mediated gene transfer.
Lab Invest
.
2011;
91:
54–62.
[CrossRef] [PubMed]
Zheng
X
de Paiva
CS
Li
DQ
Farley
WJ
Pflugfelder
SC.
Desiccating stress promotes Th17 differentiation by ocular surface tissues through a dendritic cell-mediated pathway.
Invest Ophthalmol Vis Sci
.
2010;
51:
3083–3091.
[CrossRef] [PubMed]
De Paiva
CS
Chotikavanich
S
Pangelinan
SB
IL-17 disrupts corneal barrier following desiccating stress.
Mucosal Immunol
.
2009;
2:
243–253.
[CrossRef] [PubMed]
Chauhan
SK
Dana
R.
Role of Th17 cells in the immunopathogenesis of dry eye disease.
Mucosal Immunol
.
2009;
2:
375–376.
[CrossRef] [PubMed]
Barabino
S
Rolando
M
Chen
L
Dana
MR.
Exposure to a dry environment induces strain-specific responses in mice.
Exp Eye Res
.
2007;
84:
973–977.
[CrossRef] [PubMed]
Massingale
ML
Li
X
Vallabhajosyula
M
Chen
D
Wei
Y
Asbell
PA.
Analysis of inflammatory cytokines in the tears of dry eye patients.
Cornea
.
2009;
28:
1023–1027.
[CrossRef] [PubMed]
Carreno
E
Enriquez-de-Salamanca
A
Teson
M
Cytokine and chemokine levels in tears from healthy subjects.
Acta Ophthalmol
.
2010;
88:
e250–e258.
[CrossRef] [PubMed]
Narayanan
S
Manning
J
Proske
R
McDermott
AM.
Effect of hyperosmolality on beta-defensin gene expression by human corneal epithelial cells.
Cornea
.
2006;
25:
1063–1068.
[CrossRef] [PubMed]
LaFrance
MW
Kehinde
LE
Fullard
RJ.
Multiple cytokine analysis in human tears: an optimized procedure for cytometric bead-based assay.
Curr Eye Res
.
2008;
33:
525–544.
[CrossRef] [PubMed]
Uchino
E
Sonoda
S
Kinukawa
N
Sakamoto
T.
Alteration pattern of tear cytokines during the course of a day: diurnal rhythm analyzed by multicytokine assay.
Cytokine
.
2006;
33:
36–40.
[CrossRef] [PubMed]
Narayanan
S
Miller
WL
McDermott
AM.
Conjunctival cytokine expression in symptomatic moderate dry eye subjects.
Invest Ophthalmol Vis Sci
.
2006;
47:
2445–2450.
[CrossRef] [PubMed]
Sonoda
S
Uchino
E
Nakao
K
Sakamoto
T.
Inflammatory cytokine of basal and reflex tears analysed by multicytokine assay.
Br J Ophthalmol
.
2006;
90:
120–122.
[CrossRef] [PubMed]
Cook
EB.
Tear cytokines in acute and chronic ocular allergic inflammation.
Curr Opin Allergy Clin Immunol
.
2004;
4:
441–445.
[CrossRef] [PubMed]
Cook
EB
Stahl
JL
Lowe
L
Simultaneous measurement of six cytokines in a single sample of human tears using microparticle-based flow cytometry: allergics vs. non-allergics.
J Immunol Methods
.
2001;
254:
109–118.
[CrossRef] [PubMed]
Uchino
E
Sonoda
S
Nakao
K
Sakamoto
T.
Alteration of tear cytokine balance by eye closure: analysis by multicytokine assay.
Graefes Arch Clin Exp Ophthalmol
.
2006;
244:
747–749.
[CrossRef] [PubMed]
Leonardi
A
Curnow
SJ
Zhan
H
Calder
VL.
Multiple cytokines in human tear specimens in seasonal and chronic allergic eye disease and in conjunctival fibroblast cultures.
Clin Exp Allergy
.
2006;
36:
777–784.
[CrossRef] [PubMed]
Leonardi
A
Motterle
L
Bortolotti
M.
Allergy and the eye.
Clin Exp Immunol
.
2008;
153
(suppl 1):
17–21.
[CrossRef] [PubMed]
Malvitte
L
Montange
T
Vejux
A
Measurement of inflammatory cytokines by multicytokine assay in tears of patients with glaucoma topically treated with chronic drugs.
Br J Ophthalmol
.
2007;
91:
29–32.
[CrossRef] [PubMed]
Rummenie
VT
Matsumoto
Y
Dogru
M
Tear cytokine and ocular surface alterations following brief passive cigarette smoke exposure.
Cytokine
.
2008;
43:
200–208.
[CrossRef] [PubMed]
Ward
SK
Wakamatsu
TH
Dogru
M
The role of oxidative stress and inflammation in conjunctivochalasis.
Invest Ophthalmol Vis Sci
.
2010;
51:
1994–2002.
[CrossRef] [PubMed]
Wakamatsu
TH
Dogru
M
Ayako
I
Evaluation of lipid oxidative stress status and inflammation in atopic ocular surface disease.
Mol Vis
.
2010;
16:
2465–2475.
[PubMed]
Leonardi
A
Tavolato
M
Curnow
SJ
Fregona
IA
Violato
D
Alio
JL.
Cytokine and chemokine levels in tears and in corneal fibroblast cultures before and after excimer laser treatment.
J Cataract Refract Surg
.
2009;
35:
240–247.
[CrossRef] [PubMed]
Chong
RS
Jiang
YZ
Boey
PY
Tear cytokine profile in medicated glaucoma patients: effect of monocyte chemoattractant protein 1 on early posttrabeculectomy outcome.
Ophthalmology
.
2010;
117:
2353–2358.
[CrossRef] [PubMed]
Haque
RM
Torkildsen
GL
Brubaker
K
Multicenter open-label study evaluating the efficacy of azithromycin ophthalmic solution 1% on the signs and symptoms of subjects with blepharitis.
Cornea
.
2010;
29:
871–877.
[CrossRef] [PubMed]
de Jager
W
te Velthuis
H
Prakken
BJ
Kuis
W
Rijkers
GT.
Simultaneous detection of 15 human cytokines in a single sample of stimulated peripheral blood mononuclear cells.
Clin Diagn Lab Immunol
.
2003;
10:
133–139.
[PubMed]
Ricker
LJ
Kijlstra
A
de Jager
W
Liem
AT
Hendrikse
F
La Heij
EC.
Chemokine levels in subretinal fluid obtained during scleral buckling surgery after rhegmatogenous retinal detachment.
Invest Ophthalmol Vis Sci
.
2010;
51:
4143–4150.
[CrossRef] [PubMed]
Sears
BW
Volkmer
D
Yong
S
Correlation of measurable serum markers of inflammation with lung levels following bilateral femur fracture in a rat model.
J Inflamm Res
.
2010;
2010:
105–114.
[CrossRef] [PubMed]
Yoon
KC
De Paiva
CS
Qi
H
Expression of Th-1 chemokines and chemokine receptors on the ocular surface of C57BL/6 mice: effects of desiccating stress.
Invest Ophthalmol Vis Sci
.
2007;
48:
2561–2569.
[CrossRef] [PubMed]
VanDerMeid
KR
Su
SP
Krenzer
KL
Ward
KW
Zhang
JZ.
A method to extract cytokines and matrix metalloproteinases from Schirmer strips and analyze using Luminex.
Mol Vis
.
2011;
17:
1056–1063.
[PubMed]
Huang
JF
Zhang
Y
Rittenhouse
KD
Pickering
EH
McDowell
MT.
Evaluations of tear protein markers in dry eye disease: repeatability of measurement and correlation with disease.
Invest Ophthalmol Vis Sci
.
2012;
53:
4556–4564.
[CrossRef] [PubMed]
Srinivasan
S
Menzies
K
Sorbara
L
Jones
L.
Infrared imaging of meibomian gland structure using a novel keratograph.
Optom Vis Sci
.
2012;
89:
788–794.
[CrossRef] [PubMed]
Djoba Siawaya
JF
Roberts
T
Babb
C
An evaluation of commercial fluorescent bead-based luminex cytokine assays.
PLoS One
.
2008;
3:
e2535.
[CrossRef] [PubMed]
Suzuki
M
Massingale
ML
Ye
F
Tear osmolarity as a biomarker for dry eye disease severity.
Invest Ophthalmol Vis Sci
.
2010;
51:
4557–4561.
[CrossRef] [PubMed]
Li
S
Sack
R
Vijmasi
T
Antibody protein array analysis of the tear film cytokines.
Optom Vis Sci
.
2008;
85:
653–660.
[CrossRef] [PubMed]
Thakur
A
Willcox
MD.
Cytokine and lipid inflammatory mediator profile of human tears during contact lens associated inflammatory diseases.
Exp Eye Res
.
1998;
67:
9–19.
[CrossRef] [PubMed]
Solomon
A
Dursun
D
Liu
Z
Xie
Y
Macri
A
Pflugfelder
SC.
Pro- and anti-inflammatory forms of interleukin-1 in the tear fluid and conjunctiva of patients with dry-eye disease.
Invest Ophthalmol Vis Sci
.
2001;
42:
2283–2292.
[PubMed]
Sack
R
Conradi
L
Beaton
A
Sathe
S
McNamara
N
Leonardi
A.
Antibody array characterization of inflammatory mediators in allergic and normal tears in the open and closed eye environments.
Exp Eye Res
.
2007;
85:
528–538.
[CrossRef] [PubMed]
Li
M
Cushing
SK
Zhang
J
Three-dimensional hierarchical plasmonic nano-architecture enhanced surface-enhanced Raman scattering immunosensor for cancer biomarker detection in blood plasma.
ACS Nano
.
2013;
7:
4967–4976.
[CrossRef] [PubMed]
Han
JH
Kim
HJ
Sudheendra
L
Electrophoretic build-up of multi nanoparticle array for a highly sensitive immunoassay.
Biosens Bioelectron
.
2013;
41:
302–308.
[CrossRef] [PubMed]
Inic-Kanada
A
Nussbaumer
A
Montanaro
J
Comparison of ophthalmic sponges and extraction buffers for quantifying cytokine profiles in tears using Luminex technology.
Mol Vis
.
2012;
18:
2717–2725.
[PubMed]