Influencia de los factores ambientales en los tests diagnósticos del síndrome de ojo secoestudio en una cámara de ambiente controlado

Resumen

INTRODUCTION Dry eye is recognized as a multifactorial inflammatory disorder of the lacrimal functional unit (LFU) composed by the cornea, conjunctiva, accessory lacrimal glands, meibomian glands, main lacrimal gland, and the interconnecting innervation that integrates the activities of each component (Stern et al., 1998). In DED, the interactions of the LFU components become disrupted. It is a common disorder affecting between 5.5% (McCarty et al., 1998) and 33.7% (Lin et al., 2003) of the population, depending on the criteria used for diagnosis. It is characterized by ocular discomfort and pain, visual disturbance, tear film instability, and increased tear osmolarity (DEWS, 2007a), and inflammation of the LFU. Currently, except for Cyclosporine A (CsA), there are no pharmacological therapies approved for the management of DED. Available treatments can decrease symptoms and signs, but have limited effect on the underlying cause of the disease. A clear example is artificial tears attempting to mimic human tears with physico-chemical properties (pH, osmolarity, viscosity, surface tension), and chemical elements (mainly ions) of their composing (sodium, potassium, phosphate, zinc, etc.). However, human tears have a number of more complex compounds which are impossible to mimic (antibodies, enzymes, growth factors, nutrients, etc.). Clinical tests such as tear break-up time (T-BUT), Schirmer test, the phenol red thread test (PRTT) or vital stainings (fluorescein staining, rose bengal or lissamine green) and diverse questionnaires of DED (SANDE I and II, Dry Eye McMonnies questionnaire, DEQ, OSDI© ...) are daily used to assess clinical signs and symptoms of DED, classify the disease and provide the most appropriate treatment. However, many of them have shown low reliability and reproducibility (Nichols et al., 2004). Additionally, most studies found poor or no correlation between the clinical signs and symptoms in patients with DED (Hay et al., 1998; Schein et al., 1997; Bjerrum., 1996, Moore et al., 2009, Cuevas et al., 2012) and clinical evidence indicates that the available diagnostic tests do not always correlate well with symptoms (Begley y cols., 2003), especially in mild to moderate forms of DED (Cuevas et al., 2012). During recent years, new techniques have been developed for the evaluation of both the ocular surface and the tear film, such as tear clearance test, study of growth factors and inflammatory molecules (cytokines/chemokines) in epithelial cells and tears (Lambiase et al., 2011; Enriquez-de-Salamanca et al., 2010). Some of these evaluation techniques can potentially be useful both to improve the correlation between signs and symptoms and the diagnosis of DED or other diseases of the ocular surface (DEWS, 2007d). The difficulty in demonstrating the correlation between clinical improvement of objective and subjective symptoms makes both the classification and diagnosis of DED as well as finding appropriate treatments, a complex aim to be reached. One of the most important factors implicated in this disconnection is the environment to which the LFU is exposed and which can abruptly vary. This variability is actually a well-recognized problem in the confirmation of therapeutic efficacy in the context of clinical trials (Foulks., 2003). It is reported that DED symptoms vary depending on the indoor (air conditioned rooms) and outdoor (arid regions) environmental conditions (Wolkoff., 2012). In addition, millions of individuals are exposed to numerous artificially controlled environments in their daily life, such as air-conditioned or heated rooms, vehicles, and/or airplane cabins. Thus, the LFU is exposed to multiple environmental variables including temperature, airflow, humidity, and in some cases, atmospheric pressure. These environmental variables could stimulate or retard tear film evaporation and increase or decrease symptoms and signs (Barabino et al., 2002; Rolando et al., 1998; González-García et al., 2007). The control of the environmental conditions to which patients are exposed before being evaluated might minimize the influence that these have on the LFU, which may allow the obtaining of more reliable diagnostic test values. Similarly, the ability to "homogenize" patients before being evaluated to minimize the influence of the environment could be a crucial factor to improve the reliability and eventual success of therapeutical multicenter clinical trials, where each center is located in a different geographic location with different climatic characteristics. In order to control the environment in which DED patients are examined, environmental chambers that create a controlled adverse environment are being used (Abelson et al., 2002; Ousler et al., 2002; Maruyama et al., 2004; Morgan et al., 2004, González-García et al., 2007). The use of environmental chambers could be helpful in improving the design of clinical trials and studying the effects of DED therapeutics and contact lenses (Abelson et al., 2002; Ousler et al., 2002). The Ocular Surface Group at IOBA, University of Valladolid is highly involved in the study of inflammatory diseases affecting the ocular surface, DED being its main focus. About 10 years ago, a research program in inflammation of ocular surface was started and co-directed by Prof. Margarita Calonge (IOBA, University of Valladolid) and Dr. Michael E. Stern (Allergan Inc). In the context of this program, a study with an environmental chamber located at the School of Architecture in the University of Valladolid was designed. Although this chamber was not adapted to be used with humans, but for studying the quality of air, a first study with humans was possible by performing some reformations (González-García et al., 2007). This study demonstrated the utility and the importance of this controlled environment. Thus, in 2008, an environmental chamber was constructed, becoming part of the so-called Controlled Environment Research Laboratory (CERLab). Within this chamber, environmental variables such as temperature, relative humidity (RH), and atmospheric pressure can be controlled. This doctoral thesis is the first scientific study conducted within CERLab, to evaluate the influence of environmental conditions on both DED patients and healthy subjects. HYPOTHESIS Adverse environmental conditions recreated within a controlled environmental chamber can cause a deterioration of the LFU in both DED patients and healthy individuals, as evaluated by the alteration of clinical diagnostic tests of DED and/or concentrations of various tear molecules. OBJECTIVES General objective To evaluate the effects induced by various environmental conditions recreated in the controlled environment chamber (CERLab) in DED diagnostic tests (symptoms, signs, and tear molecules) in three study samples: control subjects, patients with mild-moderate DED, and patients with severe DED. Specific objectives 1. To assess the effect of an environmental condition frequently found in Valladolid, Spain on the LFU of DED patients and healthy subjects, evaluated by the variation of clinical diagnostic tests of DED and tear molecule concentrations. The specific environmental conditions recreated in the environmental chamber were: 23ºC of temperature, 45% RH, and 930 mb of atmospheric pressure (corresponding to the one in Valladolid which is 690 m above sea level). 2. To assess the effect of an environmental condition frequently found in drier areas or in modern buildings (where RH is typically lower) on the LFU of DED patients and healthy subjects, evaluated by the variation of clinical diagnostic tests of DED and tear molecule concentrations. The specific environmental conditions recreated in the environmental chamber were the same as above but with a lower humidity: 5% RH. 3. To assess the effect of the environmental condition frequently found in air cabins during in-flight conditions on the LFU of DED patients and healthy subjects, evaluated by the variation of clinical diagnostic tests of DED and tear molecule concentrations. This specific environmental condition recreated in the environmental chamber was: 23ºC, 5% RH and 750 mb of atmospheric pressure (corresponding to 2500 m above sea level). 4. To evaluate the influence on the LFU induced by implementation of air flow on the three environmental conditions detailed above. 5. To assess the differences on DED clinical diagnostic test results in two homogeneous samples of healthy subjects living in two different geographic localizations with different climate: Valladolid (Spain), where the weather is mostly continental, and Braga (Portugal), where the climate is maritime. COMPARATIVE ANALYSIS STUDIES I AND II METHODOLOGY Participants The study protocols were approved by the institutional review board of the Institute for Applied OphthalmoBiology (IOBA), University of Valladolid and the University of Valladolid Medical School Ethics Committee. The study adhered to the tenets of the Declaration of Helsinki. All enrolled subjects were informed of the nature of the study and consent forms were signed. The study was double-masked and thus neither the participants nor the examiners knew to which of the specific randomized environmental conditions the participants were exposed. The same examiner always performed the clinical subjective evaluation of corneal and conjunctival staining, conjunctival hyperemia, and fluorescein tear break-up time (T-BUT). DED patients were recruited among known symptomatic patients already being cared for at the Ocular Immunology Unit of IOBA and Clinic Hospital, University of Valladolid. Healthy age-matched volunteers were also recruited to serve as control subjects. All of the participants were divided into two groups. The first one was assigned to the so-called ¿Standard Condition¿ and the second one to the ¿Adverse Condition¿ (see below). We selected mild to moderate (level 1 and 2) DED patients as classified by the International Dry Eye Workshop (DEWS) dry eye severity grading scheme (DEWS, 2007a). During a preliminary visit, potential patients and subjects were screened for inclusion criteria. For DED patients, these included an ocular surface disease index (OSDI©) score (Schiffman et al., 2000) > 12 and corneal fluorescein staining ¿ grade 2 (Oxford Scale) (Bron et al., 2003). For healthy control subjects, the inclusion criteria were an OSDI© score < 12 and corneal fluorescein staining ¿ grade 1 (DEWS, 2007a). In addition, individuals from both groups had to be within normal limits in at least two out of the following four tests: fluorescein T-BUT > 7 sec (Lamp et al., 2009), conjunctival lissamine green staining (Oxford Scale) ¿ grade 1 (DEWS, 2007a; Bron et al., 2003; Xu et al., 1995), Schirmer test without anesthesia ¿ 5 mm in 5 min (van Bijsterveld., 1969; DEWS, 2007a), and phenol red thread test (PRTT) > 20 mm in 15 seconds (Patel et al., 1998; Sakamoto et al., 1993) Exclusion criteria for both groups were age < 40 years old, contact lens wear, pregnancy or nursing, history of ocular surgery within the last 6 months, any acute or chronic ocular disease including patients with concomitant allergies (even if mild) other than DED, and any systemic anomaly that contraindicated being subjected to any environmental controlled condition. Those individuals who were under any medication, either ocular or systemic, had to keep it unchanged, with the same schedule during the preceding two months and for the entire duration of the study. DED patients and control subjects were instructed not to instill any eye drop (artificial tears included) within the 4 hours prior to each visit. Only one eye of each individual was included in the study. The selection was performed during the screening visit and depended first on corneal staining. The eye having more severe corneal staining was selected for the DED group. For the control group, the eye with the least corneal staining was selected. If both eyes had the same corneal staining, then the most symptomatic eye was selected for the DED group, whereas for the control group, the least symptomatic one was chosen. If the latter criteria failed to select the study eye, the eye was finally selected following a computer-generated random table. Environmental Chamber Individuals were exposed to two different environment conditions within an environmental chamber (VISIÓN I+D, SL; Valladolid, Spain) inside the Controlled Environmental Research Laboratory (CERLab), located at the IOBA building at the University of Valladolid, Valladolid, Spain. This facility was composed of an exposure chamber with a maximum capacity of 8 subjects and an evaluation chamber (Figure 1). Both chambers were 6 × 3 meters (18 m2). The following environmental conditions could be controlled in both rooms simultaneously: temperature [range: 15-30 degrees Celsius (ºC), 1ºC steps] and RH, (range: 5¿80%, 1% steps). Additionally, airflow (blower exit velocity: range 0.60¿3.60 m/s), illumination (range, 10¿1000 lux, 1 lux steps), and atmospheric pressure [range: 930¿450 millibars (mb), 1 mb steps] could also be controlled. The evaluation chamber was completely equipped with the clinical ophthalmic instruments needed to evaluate the LFU. Using a control display located outside of the chamber, environmental conditions within the chamber were monitored throughout the entire duration of the experiments and were recorded in 5 min intervals. Between 2 and 15 days after the screening visit, participants were assessed within the evaluation chamber before and after a two hour exposure to the environmental condition. Experimental sessions were run only when an anesthesiologist working at the surgery facilities (located on the same floor) was also available to take care of any possible medically adverse event. Figure 1. Environmental Chamber of the Controlled Environmental Research Laboratory (CERLab). A: engine room; B: exposure room; C: evaluation room. Environmental Conditions For two hours, each subject was exposed to either the Standard Condition or the Adverse Condition within the environmental chamber maintained at 23ºC. For the Standard Condition, the RH was 45% and the barometric pressure was 930 mb, the average atmospheric pressure found in Valladolid, which is 690 meters above sea level. For this condition, there was no localized air flow, and participants performed near vision tasks such as reading, playing cards, working crossword puzzles, doing board games, etc., for the whole exposure period. For the Adverse Condition, the RH was 5% and the atmospheric pressure was 750 mb, similar to that usually found within an aircraft cabin at 2.500 m above sea level (Federal Aviation Administration, 1996). Pressure changes were designed to simulate an actual aircraft flight, with a 15 min ascent during which the pressure decreased at a rate of 12.7 mb/min, an 85 min cruise, and a 20 min descent during which the pressured increased at a rate of 9.5 mb/min. The Adverse Condition included localized air flow with a mean velocity of 0.43 m/s, provided by individual blowers (Belnor Engineering Inc., Ottawa, ON, Canada) located one meter away from each subject. During the exposure, participants watched a documentary on a conventional LED TV monitor of 1.4 m that was situated above eye level so that patients looked slightly upwards. These conditions were designed to achieve a larger LFU area exposed to the desiccating conditions. Examination Procedure The examinations were performed in the sequence outlined below, with a 2-5 min interval between each test. Modified Single-Item Score Dry Eye Questionnaire (SIDEQ). To evaluate DED symptoms, we used a modified SIDEQ (Invest Ophthalmol Vis Sci 44 [Suppl]:2448, 2003). SIDEQ assesses the subject¿s ocular discomfort due to symptoms of dryness including dryness, sandy or gritty feeling, burning or stinging, pain, itching, sensitivity to light, and blurred vision. The level of discomfort ranging from ¿none¿ to ¿severe¿ is normally scored on a 0 to 4 scale. For this study, we assessed the same symptoms as the original one (Invest Ophthalmol Vis Sci 44 [Suppl]:2448, 2003); however, the level of discomfort was scored using a visual analogue scale (VAS) (Aitken., 1969) to increase test sensitivity. The VAS consisted of a horizontal line 100 mm long. At the left end of the line the number ¿0¿ indicated the absence of symptoms. At the right end of the line, the number ¿100¿ indicated severe symptoms. Participants performed the modified SIDEQ before beginning the chamber exposure, one hour after the beginning, and at the end of the two hour exposure. Tear osmolarity. Tear osmolarity was assessed using the TearLab (TearLab Corporation, San Diego, CA, USA). This system is designed to collect and analyze tear osmolarity from a 50 nl tear sample. Collection was performed at the external canthus in a non-traumatic procedure to avoid reflex tear secretion. A disposable chip was used for each sample (Benelli et al., 2010). Prior to proceeding with any measurement, Tearlab calibration was performed following the manufacturer¿s instructions. We used the recommended cut-off value of 316 mOsm/l for DED diagnosis (DEWS, 2007a; Tomlinson et al., 2006). Phenol red thread test (PRTT). The PRTT (Zone Quick Test; Menicon Company Ltd., Nagoya, Japan) was placed in the recommended position over the external canthus, and the length of the wetted thread was read 15 seconds later (Hamano et al., 1983). Values of 20 mm or below were considered abnormal (Patel et al., 1998). Conjunctival hyperemia. Bulbar hyperemia was scored based on the Efron scale (Efron., 1998). Nasal and temporal areas were assessed independently; however, the final score was obtained after averaging both values. Tear sample collection. Tear samples (4 µl) were obtained for cytokine/chemokine and matrix metalloproteinase (MMP)-9 analyses before instilling any vital staining. As previously described (Carreño et al., 2010), tears were collected from the external canthus in a non-traumatic manner to avoid as much as possible reflex tear secretion. The samples were collected using a glass capillary tube (Drummond Scientific Co, Broomall, PA, USA) and then diluted 1:10 by delivery with a Pipetboy (Integra Biosciences AG, Zizers, Switzerland) in a 0.5 ml microtube (Sarstedt AG&Co, Nümbrecht, Germany) containing Cytokine Assay Buffer (Merck Millipore, Millipore Iberia, Madrid, Spain). The samples were then frozen at -80°C until analysis. Fluorescein T-BUT. T-BUT was performed after instillation of 5 µl of 2% sodium fluorescein (Colircusí Fluoresceína 2%, Alcon Cusí, SA, Barcelona, Spain) into the inferior fornix with a micropipette (Capp Single Pipette EcoPipet Red, Capp Pipettes, Everton Park, QLD, Australia). The subjects were then instructed to blink several times for a few seconds to ensure adequate mixing of the dye. The tear film was viewed through a yellow Wratten #12 filter (Eastman Company, Rochester, NY, USA), and the T-BUT was taken as the interval between the last complete blink and the appearance of the first corneal black spot. The measurements were made three times, and the mean value was recorded (Norn., 1969). T-BUTs of 7 seconds or less were considered abnormal (Lam et al., 2009). Corneal fluorescein staining. Corneal staining was evaluated using a slit-lamp biomicroscope (SL-D7; Topcon Corporation, Tokyo, Japan) with a cobalt-blue filter over the light source and a yellow Wratten #12 filter. Measurements were made two min after instillation of 5 µl of 2% sodium fluorescein (Colircusí Fluoresceína 2%) into the inferior fornix of the eye. Corneal fluorescein staining was graded using both the Oxford scheme (Bron., 1997) and a modified Baylor scheme Pflugfelder et al., 2004). The Oxford grading scheme included five panels (0, I, II, III, IV and V) that represented typical gradations of stain on either the cornea or conjunctiva. The original Baylor grading scheme assessed the fluorescein staining by dividing the cornea into central, superior, temporal, inferior, and nasal zones (Pflugfelder et al., 2004a, b). Grading the intensity of the staining for each zone was based on the number of dots found on a 5-point scale: no dots, 0; one to five dots, 1; six to fifteen dots, 2; sixteen to thirty dots, 3; above thirty dots, 4. Moreover, if there was one area of confluence, then a point is added to the score obtained after summing the five corneal areas. If two or more areas were confluent, then two more points are added. If there was filamentary keratitis, another 2 points are finally added. With the aim of providing a more detailed corneal staining analysis, we modified the Baylor grading scheme by splitting it into two scales based on the number of stained dots per corneal area and the severity of the global staining. Thus, we graded each corneal area following the 5-point Baylor scale, and then we graded the severity following the Baylor scheme indications regarding presence of confluence and/or filamentary keratitis. This approach resulted in two different Baylor scores. We always used a template that was incorporated to the biomicroscope viewing system to assess the five corneal zones, thus, providing the lowest possible variability when grading. Conjunctival lissamine green staining. Conjunctival staining was evaluated with lissamine green strips (GreenGlo, HUB Pharmaceuticals, LLC, Rancho Cucamonga, CA, USA). The strips were wetted with 25 µl sodium chloride (NaCl 0.9%, B/Braun, Barcelona, Spain) and then gently applied into the inferior fornix. Staining was evaluated one minute after the instillation following the Oxford scheme grading system (Bron., 1997). Schirmer test without topical anesthesia. One Schirmer sterile strip (Tearflo, HUB Pharmaceuticals, LLC) was placed in the lateral canthus of the inferior lid margin of both eyes (van Bijsterveld., 1969). The subjects were asked to maintain the eyes closed during the test. The length of wetting was measured in millimeters after 5 min. Results of 5 mm length or more were considered normal (DEWS, 2007a). Analysis of Tear Molecules Tear levels of cytokines, chemokines, and MMP-9 were determined by commercial immune-bead based assays in a Luminex IS-100 (Luminex Corporation, Austin, TX, USA). The concentrations of epidermal growth factor (EGF), CX3CL1/fractalkine, interferon (IFN)-¿, interleukin (IL)-1RA, IL-1ß, IL-2, IL-6, CXCL8/IL-8, IL-10, IL-12p70, IL-17, CXCL10/IP-10, CCL5/RANTES, tumor necrosis factor (TNF)-¿, and vascular endothelial growth factor (VEGF) were measured simultaneously with a 15-plex assay (HCYTO-60K 15X-Milliplex Millipore Iberia, Madrid, Spain). MMP-9 concentration was measured in a separate assay with a MMP-9 single-plex assay (HMMP2-55K Panel 2, Milliplex, Millipore Iberia). The samples were analyzed following the manufacturer's protocol. Briefly, 10 µl of the 1:10 diluted sample were incubated under agitation overnight at 4°C with beads coated with antibodies specific for each cytokine/chemokine or MMP-9. After washing, the beads were incubated with biotinylated human cytokine/chemokine or MMP-9 antibodies for 1 hour, followed by incubation with streptavidin-phycoerythrin for 30 min. Standard curves, obtained from samples of known concentrations of recombinant human cytokines/chemokines or MMP-9 were used to convert fluorescence units to concentration units (pg/ml). The minimum detectable concentrations (in pg/ml) for molecules analyzed were EGF, 2.7; CX3CL1/Fractalkine, 6; IFN-¿ and TNF-¿, 0.1; IL-1RA, 2.9; IL-1ß and IL-12p70, 0.4; IL-2, IL-6, and IL-10, 0.3; CXCL8/ IL-8 and IL-17, 0.2; CXCL10/IP-10, 1.2; CCL5/RANTES, 1; VEGF, 5.8; and MMP-9, 10. The data were stored and analyzed with the "Bead View Software" (Upstate-Millipore Corporation, Watford, UK). In some samples, the assayed molecule was undetectable. To include those samples in the statistical analysis, we assigned each the minimum detectable value (provided by the assay manufacturer). However, molecules that were detected in less than 30% of the population were not further analyzed. Data Analysis Data were expressed as the mean ± standard error of the mean (SEM). Statistical analyses were performed using the Statistical Package for the Social Sciences software (SPSS 18.0 for Windows; SPSS Inc., Chicago, IL, USA) and R (RDC Team., 2006) software by a licensed statistician (MEM). The U-Mann Whitney test was used for comparisons of two independent sample groups. For comparisons between tests performed prior to and after the environmental exposure, the Wilcoxon signed rank test was used for quantitative variables and McNemar test for qualitative variables. Two-sided P-values ¿0.05 were considered statistically significant. RESULTS Screening Visit For exposure to the Standard Condition, 19 control subjects (9 males; 10 females, 60.6 ± 1.7 years old) and 15 DED patients (5 males; 10 females, 60.3 ± 2.1 years old) were selected. For exposure to the Adverse Condition, 20 control subjects (6 males; 14 females, 59.3 ± 1.9 years old) were paired with 20 DED patients (6 males; 14 females, 64.7 ± 1.8 years old). The screening evaluations for subjects in both groups were conducted under both standard and adverse environmental conditions. For both conditions, the OSDI© scores for DED patients were significantly higher than for control subjects. However, for all objective measures, there were no significant differences between the two groups under either environmental condition. No patients suffered any adverse event. Clinical Tests For tear osmolarity, conjunctival staining, and the Schirmer test, there were no statistically significant differences between the control and DED groups before or after exposure to either standard or control conditions, nor did the adverse condition induce any significant changes from the pre-exposure state for these tests. The modified SIDEQ scores of the DED group were significantly higher than those of the control group after one and two hours of exposure to the Standard Condition. Under the Adverse Condition, the scores of the DED group were significantly higher prior to, at one hour, and after two hours of exposure. For the DED group, PRT test scores were significantly decreased after two hours in the Adverse Condition. Conjunctival hyperemia scores significantly increased in all individuals after both exposures. For both groups, fluorescein T-BUT significantly decreased in the Adverse Condition but not the Standard Condition. In the Adverse Condition, fluorescein T-BUT was significantly less in the DED group than the control group both before and after the exposure. Corneal staining scored by the Oxford method of assessment increased significantly for the control group in the Standard Condition and for both groups in the Adverse Condition. Likewise, using the Baylor scheme, there was an increase in the number of stained dots found in the five evaluated areas in both groups. The DED group showed significantly increased staining independently of the assessment method, Oxford or Baylor scheme, used. Analysis of Tear Molecules Among the 16 molecules analyzed in tears, CX3CL1/Fractalkine, EGF, CXCL8/IL-8, IL-1RA, CXCL10/IP-10, and MMP-9 were detected in more than 90% of both groups prior to and after undergoing either controlled environmental condition. IL-6, CCL5/RANTES, and VEGF were detected in 58-80% of both groups prior to entering the environmental chamber and 69-90% after exposure to either environment. The detection rates for IFN-¿, IL-12p70, IL-17, IL-10, IL-1ß, and TNF-¿ were below 30% for both groups under both exposure conditions, and therefore not considered to be statistically relevant. EGF, IL-6, CCL5/RANTES, and MMP-9 tear levels changed significantly after exposure to the Adverse Condition in the (Figure 2). For both the control and DED groups, EGF significantly decreased and IL-6 increased after exposure to the Adverse Condition. We also found a significant increase in CCL5/RANTES values in the control group after the Adverse Condition; however, the increase in the DED group was not significant. While MMP-9 values for both groups increased after exposure to both environmental conditions, only the increase in the DED group after the Adverse Condition was significant. Non-significant changes were detected in tear levels of all the other molecules that were analyzed. STUDY III. JUSTIFICATION MODEL HYPOTHESIS Healthy subjects from two different parts of the world with different climatic conditions will present different basal values¿ in the clinical diagnostic tests that evaluate the LFU, due to the influence of environmental conditions. To test this hypothesis, we chose a geographic location climatically different from Valladolid. Valladolid¿s climatic characteristics correspond to a cold-mediterranean (with continental-like features) weather. Braga, in the north coast of Portugal, has temperate-cold maritime climate and there is where Dr. Méijome¿s team is located (CEORLab, Department of Physics, University of Minho). OBJECTIVE The objective of this study was to evaluate a sample of healthy subjects using the same protocol previously detailed for Study I and II and compare them with a similar sample evaluated in Valladolid as part of the study I and II. METHODOLOGY Participants The study protocols were approved by the institutional review board of the Institute for Applied OphthalmoBiology (IOBA), University of Valladolid and the University of Valladolid Medical School Ethics Committee. The study adhered to the tenets of the Declaration of Helsinki. All enrolled subjects were informed of the nature of the study and consent forms were signed. The same examiner of studies I and II (MTY) always performed all evaluation tests of this study. A sample of 14 healthy controls (7 men and 7 women), belonging to the previous studies I and II performed in Valladolid was selected. This group was composed with all individuals who participated in both studies during the months of April, May, and June 2011 and 2012, as these were the same months in which the Braga study was carried out. During this period, 14 healthy individual (7 men and 7 women), age and gender-matched to those previously evaluated were screened and eventually examined in two more visits. The evaluation tests performed during the screening visit and during the 2 evaluation visits, as well as their sequence, were the same previously used in Valladolid (see comparative analysis studies I and II). Tear osmolarity measurement and analysis of molecules in tears could not be performed because the TearLab® and the Luminex machine were not available at the CEORLab and it was decided again their shipment. Only one eye of each individual was included in the study. The selection was performed with the same procedure using in the protocol of doctoral study in Valladolid (see methodology: participants of comparative analysis studies I and II). Data was analyzed with SPSS software (RDC Team., 2006). Reliability of diagnostic tests was analyzed statistically to determine the best way to compare the measurements obtained in both cities. Intraclass Correlation Coefficient (ICC) (Fleiss et al., 1986) for quantitative variables (PRTT, T-BUT and Schirmer test) and Kappa Coefficient (Landis., 1977) for ordinal scale variables (corneal and conjunctival staining and conjunctival hyperemia) were used to analyze variability of measures. Both ranges were between 0 and 1. The total variability of the measurements had two components: the variability due to differences between subjects and the variability due to differences between the measurements for each subject. The maximum possible agreement corresponding to the value 1, which means that all the variability observed is due to differences between subjects and not at the different measurements taken at different times. A 0 value means that the observed agreement is the same as it would be expected by random. RESULTS Reliability of Diagnostic Tests In general terms, all clinical tests showed a low reliability. This poor reliability was worse in Braga than in Valladolid. Due to the lack of reproducibility of clinical tests conducted in both cities, and following IOBA statistical unit advice, the data could not be analyzed using the mean of the three measurements for both cities. Therefore, we analyze only the visit that was the median of the three visits. Temperature and RH No significant differences were observed in the average temperature between Braga and Valladolid; however both cities had a significantly different RH humidity, as expected. Temperature (ºC) Valladolid 16.00 ± 0.89 0.762 Braga 16.50 ± 0.71 Relative Humidity ¿ RH (%) Valladolid 52.60 ± 3.69 0.003 Braga 81.25 ± 1.22 RH: relative humidity; SEM: standard error of the mean. Clinical test The ocular surface of the control subjects included in Braga (Portugal) showed better baseline values than those evaluated in Valladolid, as measured with the conventional DED diagnostic tests. We found significant higher values on conjunctival hyperemia (temporal) (p= 0.02), corneal staining (Oxford scale: p= 0.01; Global Baylor scale: p= 0.02; Temporal area (Baylor scale): p= 0.01) and mean conjunctival staining (p= 0.03) in Valladolid sample compared to that in Braga. No significant differences were found in PRTT and Schirmer test. Braga subjects showed a better T-BUT than Valladolid subjects (p= 0.0006). Additionally, control subjects from Braga had a significantly decreased score in the DED symtom questionnaire than subjects from Valladolid (p= 0.002). CONCLUSIONS 1. We managed to recreate six different environmental conditions within the environmental chamber of the IOBA-CERLab, all of which produced variable changes in the LFU of DED patients after a 2-hour exposure time. These environmental conditions simulated: 1) the warm-to-cold Mediterranean climate (formerly called continental weather) of Valladolid, Spain (where the IOBA-CERLab is located); 2) the indoor conditions of a conventional building, characterized by low humidity and air flow; and 3) the low humidity and high altitude of a in-flight air cabin of a commercial airplane. 2. Based on the results obtained under each of the 6 different condition tested, we could select the most adequate environmental condition or combination of them for future studies or clinical trials depending upon the specific hypothesis. Thus, we could recreate within the CERLab the following kind of environmental conditions that would presumably affect the LFU: 2.1. The outdoor climate of a specific geographic location, with the ability to recreate low humidity and/or high altitudes. 2.2. The indoor conditions similar to those found in modern buildings where acclimatization is reached through conditioned (heated or cooled) air flow and, consequently, expose individuals to a low humidity and to air flow near their faces. The rationale for this is that our study demonstrated that low humidity worsens the LFU and that further implementation of indirect air flow enhances these LFU changes. 2.3. A sudden exposure to adverse environmental conditions consisting on low humidity and air flow, simulating the scenario that DED patients suffer when entering places such as a shopping centers, movie theaters, transportation vehicles (cars, train, buses, etc) or, in many cases their work places. By recreating these conditions, we could test whether a certain therapeutic approach would ameliorate or even prevent these disease exacerbations. 2.4. An indoor condition simulating the one found in an in-flight commercial airplane cabin, which has produced LFU changes after only a 2 hour exposure time. Our findings would offer the opportunity to start a new research line regarding the assessment of DED therapeutics aimed at reducing the visual complaints of individuals during and/or after flying. 2.5. Finally, we can simulate a predetermined visual task taking into account that visual display units or TV visualization tend to provoke more remarkable LFU changes. 3. Mild to moderate DED patients subjected to controlled adverse conditions like low humidity and indirect air flow might worsened, reaching signs similar to those characterizing severe DED patients. This fact suggest that milder, instead of more severely affected patients, could be recruited to test therapeutics for severe DED. This is relevant due to the known difficulty to recruit severe patients for therapeutic clinical trials and to the low chance of improvement for these patients in the short-time duration of these trials. 4. The study performed in healthy individuals in two cities (Valladolid, Spain and Braga, Portugal), with significantly different RH, showed that baseline values for DED diagnostic tests are not comparable, thus lacking reproducibility. This finding highlights the relevance of standardizing the environmental conditions in multicenter clinical trials, where outcome variability is expected associated to the different climatic conditions in each geographic location. This variability can be one of the reasons why therapeutic multicenter clinical trials fail when previous unicenter trials succeeded. 5. We have defined and validated a complete DED diagnostic protocol designed to detect changes in the LFU of DED patients and healthy subjects after exposure to diverse controlled environmental conditions for 2 hours. This evaluation protocol lasts 40 minutes and consists of a DED symptom questionnaire, 8 clinical diagnostic tests and tear collection for its tear molecule concentration assessment. It can be performed in four individuals at the same using two skilled clinically-oriented researchers. This wide protocol, and based on the outcomes obtained in this study, can be shortened and tailored depending upon the specific hypothesis to demonstrate, the therapeutic mechanism of action to test and/or the type of individuals recruited (i.e. healthy subjects, moderate or severe DED patients). 6. Finally, it has been demonstrated the importance of analyzing tear molecules as potential inflammatory activity biomarkers, because the concentration of some of them (especially MMP-9) varied after the exposure to predetermined environmental conditions.