Original Paper

Ophthalmic Res 2008;40:249–256 Received: September 20, 2006 DOI: 10.1159/000127832 Accepted after revision: July 6, 2007 Published online: April 25, 2008

Increased Homocysteine Levels in Tear Fluid of Patients with Primary Open-Angle

a b a c b J.B. Roedl S. Bleich U. Schlötzer-Schrehardt N. von Ahsen J. Kornhuber a a a G.O.H. Naumann F.E. Kruse A.G.M. Jünemann a b Department of , University Hospital, and Department of Psychiatry and Psychotherapy, c Friedrich-Alexander University, Erlangen-Nuremberg , and Department of Clinical Chemistry, Georg-August University, Goettingen , Germany

Key Words Introduction Glaucoma ؒ Homocysteine ؒ Tears ؒ Vitamins The pathogenesis of primary open-angle glaucoma (POAG) is still not fully understood. There is accumulat- Abstract ing evidence that vascular [1] , extracellular matrix [2] , Aims: We assessed homocysteine (Hcy) levels in tear fluid and neurotoxic [3] changes contribute to the develop- and plasma of patients with primary open-angle glaucoma ment of POAG besides increased (POAG). We determined the association between Hcy levels, (IOP). Extracellular matrix and proinflammatory chang- and B vitamin status. Methods: This pro- es are also involved in ocular surface disorders seen in spective case-control study included 36 patients with POAG POAG including fibrosis of the filtering bleb after tra- and 36 controls. Hcy concentrations were measured by high- beculectomy [4, 5] and dry eye syndrome [6, 7] . performance liquid chromatography. Results: Patients with Homocysteine (Hcy) is a highly reactive amino acid POAG had significantly higher mean Hcy levels both in tear that can induce typical pathological changes found in fluid (205 8 84 nmol/l; p ! 0.001, t test) and in plasma (13.43 POAG including apoptotic death of retinal ganglion cells 8 3.53 ␮mol/l; p = 0.001, t test) than control subjects (130 8 [3] , vascular [8–10] , extracellular matrix [8, 11, 12] , and 53 nmol/l and 10.50 8 3.33 ␮mol/l, respectively). Hcy in tear proinflammatory alterations [13] . In POAG and pseudo- fluid was significantly correlated with plasma Hcy in POAG exfoliation glaucoma (PEXG), elevated plasma Hcy levels patients (r = 0.459; p = 0.005, Pearson’s correlation), but not were observed in previous studies of our group [14–17] . in controls (r = 0.068; p = 0.695). POAG patients with dry eye Moreover, increased Hcy values in aqueous humor were disease had significantly higher Hcy levels both in tear fluid found both in POAG [17] and in PEXG [16] . Accordingly, and plasma than POAG patients without dry . we detected a significantly higher prevalence of the most There was no association between Hcy levels and B vitamin common genetic cause for hyperhomocysteinemia in status in subjects with POAG. Conclusions: The study sug- POAG [18]. Several following studies confirmed our re- gests increased Hcy levels in tear fluid and plasma of pa- sults of increased Hcy levels in PEXG [19, 20]. Hyperho- tients with POAG. Elevated Hcy levels might be a risk factor mocysteinemia was also shown to be a risk factor for var- for POAG and dry eye syndrome in subjects with glaucoma. ious other ocular diseases [21–23] . However, there are in- Copyright © 2008 S. Karger AG, Basel

© 2008 S. Karger AG, Basel Johannes B. Roedl, MD, Department of Ophthalmology 0030–3747/08/0405–0249$24.50/0 Friedrich-Alexander University of Erlangen-Nuremberg Fax +41 61 306 12 34 Schwabachanlage 6, DE–91054 Erlangen (Germany) E-Mail [email protected] Accessible online at: Tel. +49 9131 853 4519, Fax +49 9131 853 6401 www.karger.com www.karger.com/ore E-Mail [email protected] Table 1. Characteristics of POAG patients and controls Nongenetic determinants of Hcy Controls POAG p value (n = 36) (n = 36) (differences)

Age (mean 8 SD), years 68.589.8 67.388.2 0.59a Male sex 17 (48) 15 (42) 0.81b Caffeinated beverages (>3 per day) 13 (36) 12 (33) 1.0b Alcoholic drinks (>1 per day) 5 (14) 7 (20) 0.54b Smoking 6 (17) 8 (22) 0.77b Systemic hypertension 12 (33) 14 (39) 0.81b mellitus 4 (11) 5 (14) 1.0b Hyperlipidemia 6 (17) 4 (11) 0.74b Atherosclerotic vascular disease 4 (11) 3 (8) 1.0b

Figures in parentheses are percentages. a t test; b Fisher’s exact test.

consistent results in the literature regarding plasma Hcy mm Hg), and normal eye examination, including open angles, levels of POAG patients [14, 17, 24–27] , possibly due to normal appearance of the optic disks and normal visual fields. Each patient was evaluated for dry eye syndrome based on three different assessments of critical determinants of Hcy, criteria which had been used in previous studies [28, 29]: (1) his- such as nutritional and B vitamin status of enrolled sub- tory of, treatment of, or symptoms of dry eye syndrome defined jects. In order to illuminate the role of Hcy in glaucoma- according to the Ocular Surface Disease Index [30] questionnaire; tous neuropathy and its possible implication in ocular (2) tear dynamics which were assessed by the Schirmer 1 test and surface disorders associated with POAG, we evaluated as- tear breakup time; if one of those two tests was positive (Schirm- er 1 test ! 10 mm, or breakup time ! 10 s), the tear dynamics were sociations between Hcy in plasma, Hcy in tear fluid, and considered abnormal; (3) ocular surface abnormalities which blood B vitamins. were identified by positive vital staining with fluorescein. Pa- tients with positive results in all three criteria were considered to have definite dry eye, whereas patients with abnormalities in only two of the three criteria were defined to have probable dry eye Materials and Methods disease [28] . Exclusion criteria for controls and POAG patients were: medical history of disorders associated with increased Hcy Patients and Design levels such as thromboembolic, renal, hepatic, gastrointestinal or In this investigation, 36 consecutive Caucasian German pa- neurologic disease, as well as prior ocular surgery, a history of tients with POAG and 36 healthy controls without glaucoma, but ocular inflammation, age-related , diabetic concurrent diagnosis of were enrolled. All subjects un- , , retinal occlusive disease, rubeosis iridis and derwent glaucoma or cataract surgery at the Department of Oph- glaucoma other than POAG. Patients taking hormone substitu- thalmology at the University of Erlangen-Nuremberg, Germany. tion and medications known to affect Hcy measurements were Patients of the present clinical trial were not enrolled in any of our excluded. previous studies [14–17] . Informed consent was obtained from all Matching of controls and POAG patients was performed by participants after detailed explanation of the purpose and meth- demographic, common clinical, and lifestyle factors known to ods of the trial. The University of Erlangen-Nuremberg Human cause increased Hcy levels (table 1 ) [31] . Subjects Committee approved all protocols of this prospective Nutritional and lifestyle status was assessed using the Sub- case-control study, and the methods described adhered to the te- jective Global Nutritional Assessment test [32] . Since all pa- nets of the Declaration of Helsinki. All controls and patients were tients and controls were classified as well nourished, we used 11.7 thoroughly examined by slit lamp inspection, applanation to- ␮ mol/l as the primary cutoff value to define hyperhomocystein- nometry, fundoscopy, gonioscopy, pupillometry, and perimetry emia as in previous investigations [14, 16] . Additionally, we per- (Octopus 101, program G1). All POAG patients had open angles formed statistical analysis with 15.0 ␮ mol/l as the threshold. on gonioscopy, characteristic glaucomatous optic disk damage Twenty-nine of the 36 patients in the POAG group were on topical (excavation, presence of neuroretinal rim thinning, notching, or antiglaucomatous therapy: 53% (19/36) of the POAG patients were retinal nerve fiber layer defects), and pathological cumulative on beta-blocker, 47% (17/36) on prostaglandin analog, 36% (13/36) perimetric defect curves, that is, local as well as diffuse visual field on carbonic anhydrase inhibitor, 19% (7/36) on alpha-agonist, 3% loss in white-on-white perimetry. In POAG patients, the mean (1/36) on cholinergic, and 3% (1/36) on adrenergic topical medica- duration of disease was 11.2 8 5.2 years, the mean IOP was 18.3 tion. In order to avoid the possible influence of eyedrops on Hcy 8 2.7 mm Hg, and the global indices mean defect was 11.7 8 2.5 measurements, topical therapy was stopped in all patients at least dB. Control subjects had no history of ocular disease (except re- 12 h prior to sampling of reflex tears. fractive error, , or ), normal IOP values (! 21

250 Ophthalmic Res 2008;40:249–256 Roedl et al.

L a b o r a t o r y A n a l y s e s All patients underwent blood and tear fluid collection at the 500 25 day of ocular surgery, ensuring that the patients were fasting. Se- cretion of reflex tears was incited by the yawn reflex and by stim- 400 20 ulation of the trigeminal nerve with a nasal alcohol stimulus. A disposable glass capillary was placed near the cul-de-sac of the eye and 100 ␮ l of tear fluid were collected according to Choy et al. 300 15 [33] . Reflex tear collection with the capillary tube was shown to be the method of choice in the study by Choy et al. [33] . They com- pared basal tears with reflex tears as sources of tear fluid, and 200 10 Schirmer strips with capillary tubes as collection methods. The

Plasma Hcy (µmol/l) Schirmer strips have the disadvantage that they are associated Lacrimal fluid Hcy (nmol/l) 100 5 with direct conjunctival contact and therefore risk of vascular transudation and cell damage with leakage of plasma and intra- cellular constituents [34, 35]. Hcy is found both in plasma and in 0 0 cells and Schirmer strips might therefore lead to spuriously high Controls POAG Controls POAG levels of Hcy in tears. The use of a capillary tube for tear collection a (n = 36) (n = 36) b (n = 36) (n = 36) is much less invasive [34], allows collection of larger volumes of tear fluid than with the Schirmer strip [35], and is therefore the suggested method of choice for tear collection. There is currently Fig. 1. Distribution of Hcy levels in tear fluid ( a) and plasma ( b ) no suitable method available for collection of basal tears for bio- for controls and POAG patients. The box length is the interquar- chemical analysis [33] . tile range with the median represented by the horizontal line. Probes were immediately placed on ice, promptly centrifuged y = Case with more than 1.5 box lengths from the upper edge of (2,000 g , 10 min) within 30 min after collection and stored at the box.

–20 ° C until biochemical analysis was performed. For analysis of Hcy in tears, 50 ␮l of undiluted tear fluid were mixed with 14 ␮l of reduction agent (1.43 mol/l sodium borohy- dride). The samples were then gently vortex-mixed and incubated for 5 min at room temperature. We added 50 ␮l of precipitation plasma, Hcy in tear fluid, age, and B vitamins was assessed using agent (1.50 mol/l perchloric acid) and vortex mixed the samples Pearson’s correlation coefficient. In controls and patients, uni- for 30 s. After removal of protein with centrifugation (12,000 g , variate analysis followed by multiple regression analysis was ex- 5 min), 40 ␮l of the probes were mixed with 80 ␮ l of derivatiza- ecuted to find nongenetic determinants of Hcy. After adjustment tion agent (10.0 mmol/l monobromobimane). The tubes were for confounding factors, logistic regression analysis was done to calculate odds ratios for POAG associated with elevated Hcy capped, mixed, and incubated at 42 ° C. The autosampler (model: Midas, Spark Systems € , Emmen, The Netherlands) injected 50 ␮ l levels. aliquots onto a 4.6 ! 150 mm LC8 Supelcosil column (Supelco€ ,

Bellefonte, Pa., USA) with a column temperature of 25 ° C. For de- termination of homocysteine-S-bimane, the column was devel- R e s u l t s oped at a flow rate of 1.2 ml/min. Homocysteine-bimane adduct was detected fluorometrically with an RF-10Axl high-perfor- mance liquid chromatography detector (Shimadzu Corporation € , Hcy was detected in all tear fluid and plasma samples. Kyoto, Japan). The excitation wavelength was set at 385 nm, and Hcy levels in tear fluid and in plasma were significantly the emission wavelength at 515 nm. Standard curves and quality higher in patients with POAG than in control patients control samples were run with each batch. (fig. 1; table 2). Elevation of Hcy both in tear fluid and in For analysis of Hcy in plasma, 20 ␮ l of the diluted plasma ␮ plasma was associated with a higher risk for POAG com- probe (1: 10) were mixed with 30 l of water before the reduction agent was added. The following steps of the analysis were identical pared to controls in a logistic regression model (table 2). to the preparation of tear fluid. In POAG , a significant correlation with a moder- V i t a m i n B 6 concentrations were analyzed by high-perfor- ate degree of strength could be observed between Hcy in € mance liquid chromatography (kit by Chromsystems , Munich, plasma and Hcy in tear fluid (r = 0.459; p = 0.005; fig. 2 a) Germany), and serum vitamin B 12 and folate concentrations were determined by immunoassay (Immulite, DPC Biermann € , Bad as opposed to control subjects in whom no significant Nauheim, Germany). correlation was found (r = 0.068; p = 0.695; fig. 2b). The association between Hcy levels in tear fluid and plasma Statistical Analysis in POAG patients was further demonstrated in a general All statistical tests were two-sided and a p value less than 0.05 linear analysis, in which only the plasma Hcy (p = 0.005) was considered statistically significant. The Kolmogorov-Smirnov test showed that Hcy, B vitamins, and age were normally distrib- was a significant predictor of tear fluid Hcy, but not age uted. Differences between patient groups were evaluated by the (p = 0.316), sex (p = 0.510), or B vitamin levels (B 12: p = Fisher exact, ANOVA, and t tests. Correlation between Hcy in 0.144, B 6 : p = 0.755, folate: p = 0.151). In controls, this as-

Homocysteine in Primary Open-Angle Ophthalmic Res 2008;40:249–256 251 Glaucoma Table 2. Hcy levels in tear fluid and plasma were significantly increased in POAG patients using t tests and displayed graded risk fac- tors for POAG in a logistic regression model

Variable Controls POAG p value Odds ratio (logistic regression) (n = 36) (n = 36) (differences)

Hcy in tear fluid, nmol/l 130853 205884 t = –4.52; OR = 1.015a p < 0.001 95% CI = 1.005–1.025 p = 0.003 Hcy in plasma, ␮mol/l 10.5083.33 13.4383.53 t = –3.63; OR = 1.356b p = 0.001 95% CI = 1.131–1.625 p = 0.001 Hyperhomocysteinemia >11.7 ␮mol/l 22% (8/36) 64% (23/36) p = 0.001 OR = 6.72c 95% CI = 2.25–20.07 p = 0.001 Hyperhomocysteinemia >15.0 ␮mol/l 8% (3/36) 39% (14/36) p = 0.004 OR = 7.80c 95% CI = 1.93–31.51 p = 0.007

a Odds ratio for POAG per 1 nmol/l elevation of Hcy in tear fluid after adjustment for age, folate, caffeine, and plasma Hcy. b Odds ratio for POAG per 1 ␮mol/l elevation of Hcy in plasma after adjustment for age, folate, and caffeine. Hyperhomocystein- emia, defined on two different cutoff levels, was more prevalent in POAG using Fisher’s exact tests and was a significant risk factor for POAG compared to controls. c Odds ratio for POAG associated with hyperhomocysteinemia.

Table 3. B vitamin and creatinine levels in controls and POAG patients Variable Controls POAG p value (n = 36) (n = 36) (differences)

Folate in serum, ng/ml 12.5486.43 11.5585.58 t = 0.70; p = 0.49

Vitamin B12 in serum, pg/ml 4258194 4388243 t = –0.27; p = 0.79

Vitamin B6 in plasma, ng/ml 13.1287.61 12.5087.15 t = 0.35; p = 0.73 Creatinine in serum, mg/dl 0.8380.18 0.8180.16 t = 0.70; p = 0.49

No significant differences were found between controls and POAG patients in the above important confounding factors of hyperhomocysteinemia.

sociation between tear fluid and plasma Hcy was not sig- We evaluated the relationship between Hcy levels and nificant (p = 0.550). major nongenetic determinants of Hcy in a general linear There was no significant difference in potential con- model for both groups. In POAG patients, neither Hcy in founding factors between POAG patients and controls: plasma nor Hcy in lacrimal fluid was significantly related levels of B vitamins and creatinine, the most important to any of the nongenetic determinants of Hcy. However, nongenetic determinants of elevated Hcy, did not differ in the control group, folate (p = 0.008) demonstrated a significantly between POAG patients and controls (ta- significant negative association, whereas age (p = 0.009) ble 3 ). and caffeine consumption (p = 0.036) showed a signifi-

252 Ophthalmic Res 2008;40:249–256 Roedl et al.

500 500

× 400 400

300 300 × × ×× × × 200 × 200 × × ×××× ××× ×× × × ××× × × × × × ×

Lacrimal fluid Hcy (nmol/l) Lacrimal fluid Hcy (nmol/l) 100 100 × × × × × × × × × × × × × 0 0 0 5 10 15 20 25 0 5 10 15 20 25 a Plasma Hcy (µmol/l) b Plasma Hcy (µmol/l)

Fig. 2. Scatterplot graphs with linear regression lines in POAG patients (n = 36) ( a ) and controls (n = 36) (b ). a Positive correlation between tear fluid and plasma Hcy (r = 0.459; p = 0.005) in POAG patients. y = Patients with topical antiglaucomatous therapy; ! = patients without topical antiglaucomatous therapy. b No signifi- cant correlation in controls (r = 0.068; p = 0.695).

cant positive association with plasma Hcy levels in the patients with dry eye syndrome (252 8 66 nmol/l vs. 156 multiple regression model. Hcy in tear fluid was nega- 8 40 nmol/l; t = –3.0, p = 0.010). Both in patients with tively associated with serum folate (p = 0.047) in the con- (n = 16) and without dry eye disease (n = 56), Hcy in tear trol group. fluid was significantly correlated with Hcy in plasma (r = Prevalence of definite dry eye syndrome was signifi- 0.509, p = 0.044, and r = 0.272, p = 0.043, respectively). cantly higher in POAG patients compared to controls Topical therapy was abandoned 12 h prior to collec- [28% (10/36) vs. 8% (3/36), respectively; p = 0.035, Fisher’s tion of reflex tears, in order to avoid contamination of exact test], whereas prevalence of probable dry eye dis- tear fluid with eyedrops and thus interference of Hcy ease was not significantly different between both groups measurements. Despite these precautions, a possible in- [31% (11/36) vs. 14% (3/36), respectively; p = 0.155, Fish- fluence of chronic antiglaucomatous medication on Hcy er’s exact test]. POAG patients with probable dry eye dis- levels could not be ruled out. Therefore, we additionally ease had significantly higher Hcy levels than subjects evaluated possible associations between topical therapy without dry eye syndrome both in tear fluid (252 8 67 prior to the 12-hour pause and Hcy levels in tear fluid and nmol/l vs. 185 8 84 nmol/l; t = –2.33, p = 0.026) and in dry eye disease. plasma (15.89 8 3.08 ␮ mol/l vs. 12.34 8 3.20 ␮ mol/l; t Tests for the various types of antiglaucomatous eye- t = –3.09, p = 0.004). In the control group, patients with a drops were performed and no significant association be- diagnosis of dry eye did not have significantly increased tween Hcy levels in tear fluid and any topical therapy Hcy levels compared to subjects without dry eye disease could be observed (t tests; p 1 0.3 in all types of eye med- in tear fluid (156 8 40 nmol/l vs. 126 8 54 nmol/l; t = ication; table 4 ). Furthermore, the 29 POAG patients with –1.18, p = 0.245) or plasma (12.86 8 6.01 ␮ mol/l vs. 10.11 and the 7 POAG patients without topical antiglaucoma- 8 2.65 ␮mol/l; t = –1.77, p = 0.368). After exclusion of all tous drugs showed no significant difference in Hcy levels POAG and control patients with dry eye disease, patients of tear fluid (201 8 77 nmol/l vs. 221 8 116 nmol/l; t = with POAG still had significantly higher Hcy levels com- 0.559, p = 0.580). Besides that, Hcy levels in tear fluid and pared with control patients both in plasma (12.34 8 3.20 the number of different topical drugs used for glaucoma ␮ mol/l vs. 10.11 8 2.65 ␮mol/l; t = –2.86, p = 0.006) and therapy were not significantly associated (x = number of tear fluid (184 8 84 nmol/l vs. 126 8 54 nmol/l; t = –3.16, different antiglaucomatous eyedrops; n = number of pa- p = 0.003). Furthermore, POAG patients with dry eye tients): x = 0 (n = 7): 221 8 116 nmol/l; x = 1 (n = 12): 212 syndrome had higher Hcy levels in tear fluid than control 8 85 nmol/l; x = 2 (n = 9): 177 8 68 nmol/l; x = 3 (n =

Homocysteine in Primary Open-Angle Ophthalmic Res 2008;40:249–256 253 Glaucoma Table 4. Association between Hcy levels in tear fluid and therapy values found in other diseases with Hcy as a risk factor with different antiglaucomatous eyedrops in POAG patients (n = including vascular [9, 10] and ocular [19–23] disorders. number of patients) However, in the study by Wang et al. [25] , who found no Yes No p value significant differences between Hcy levels in POAG and controls, the mean Hcy value of the control subjects was Beta-blockers (n = 19) (n = 17) above most thresholds for hyperhomocysteinemia. It was Hcy in tear fluid, nmol/l 199869 2128101 t = 0.431; at least 1 ␮ mol/l higher compared to controls of other in- p = 0.669 vestigations [9, 19–23] including 27 studies evaluated by Prostaglandin analogs (n = 17) (n = 19) Boushey et al. [10] in a meta-analysis. Furthermore, dif- Hcy in tear fluid, nmol/l 211885 201886 t = –0.353; ferent ethnicities of the enrolled patients and different p = 0.726 evaluation of critical nongenetic determinants of Hcy Carbonic anhydrase inhibitors (n = 13) (n = 23) [31], such as assessment of nutritional status, B vitamin Hcy in tear fluid, nmol/l 198871 210892 t = 0.401; and creatinine levels, might be responsible for the incon- p = 0.691 sistent results in the literature. Alpha-agonists (n = 7) (n = 29) Hcy in plasma showed a significant and moderately Hcy in tear fluid, nmol/l 190880 209886 t = 0.528; strong correlation with Hcy in tear fluid in POAG pa- p = 0.601 tients, but not in controls. Furthermore, POAG patients Cholinergics (n = 1) (n = 35) with dry eye syndrome had significantly higher Hcy lev- 8 Hcy in tear fluid, nmol/l 130 207 85 t = 0.903; els in tear fluid and plasma than POAG patients without p = 0.373 dry eye disease. Subjects with POAG suffered signifi- Adrenergics (n = 1) (n = 35) cantly more often from definite dry eye syndrome than 8 Hcy in tear fluid, nmol/l 180 206 85 t = 0.300; control patients. To our knowledge, this is the first p = 0.766 study evaluating the possible clinical role of Hcy in Topical therapy was stopped 12 h prior to sampling of reflex tear fluid. tears to avoid contamination of tear fluid. Chronic antiglaucoma- Hcy can induce deposition of extracellular matrix tous treatment prior to this 12-hour pause was not associated with components, fibrosis [11], oxidative stress, and produc- Hcy levels in tears (t tests; p > 0.3 in all types of topical medica- tion of those subtypes of matrix metalloproteinases [8] , tion). transforming growth factors, collagens [12] , and proin- flammatory factors [13] , which were found to be increased in dry eye syndrome [7] and insufficient filtering blebs [4, 5] . Therefore, further studies are warranted to investigate 4): 225 8 70 nmol/l; x = 4 (n = 4): 203 8 94 nmol/l; p = whether increased Hcy levels in tear fluid might be im- 0.830, F = 0.367, one-way ANOVA. No association was plicated in dry eye syndrome or failure of filtering blebs found between any of the antiglaucomatous eye medica- in POAG eyes. tions and dry eye syndrome (Fisher’s exact test: beta- The result of a previous investigation [18] , which blockers, p = 1.0; prostaglandin analogs, p = 0.281; car- showed that hyperhomocysteinemia in POAG might be bonic anhydrase inhibitors, p = 0.475; alpha-agonists, caused by an endogenous genetic variant, is indirectly p = 1.0; cholinergics, p = 0.306; adrenergics, p = 1.0). Fur- supported by our finding that elevated Hcy levels in plas- thermore, the number of different antiglaucomatous ma and tear fluid in POAG were not associated with low drugs was not associated with dry eye syndrome (p = B vitamin levels or other important nongenetic determi- 0.302; ␹2 test). nants of Hcy, such as increasing age or caffeine consump- tion, as opposed to controls. Similar associations were found for Hcy in aqueous humor and nongenetic deter- Discussion minants in a previous study [17] . The study has limitations. Due to the prospective In the current study, increased Hcy concentrations in study design and the complex nature of collecting tear both tear fluid and plasma of POAG patients were de- fluid, the study population in this investigation was rela- tected. tively small (n = 72), especially the number of patients The measured mean plasma Hcy levels in POAG pa- with dry eye syndrome (n = 16). Further studies are need- tients and controls of the present study were similar to ed with larger patient numbers to confirm the possible

254 Ophthalmic Res 2008;40:249–256 Roedl et al. association between increased Hcy levels in tear fluid and tenuate the possible negative effects of Hcy. However, dry eye syndrome in POAG patients. there are conflicting results in the literature regarding In conclusion, our results suggest that elevated Hcy plasma Hcy levels in POAG. Thus, further investigations levels in tear fluid and plasma might be potential risk fac- are necessary to elucidate the potential clinical role of tors for POAG and dry eye disease in patients with glau- Hcy in glaucoma patients, such as the possible association coma. Therapy of high Hcy levels by administration of between elevated Hcy values and the higher prevalence systemic or topical B vitamin preparations would be safe, of systemic vascular dysregulations or ocular surface dis- effective and inexpensive [10, 12] , and might therefore at- orders in POAG.

References

1 Emre M, Orgul S, Gugleta K, Flammer J: Oc- 10 Boushey CJ, Beresford SA, Omenn GS, Motul- 19 Vessani RM, Ritch R, Liebmann JM, Jofe M: ular blood flow alteration in glaucoma is re- sky AG: A quantitative assessment of plasma Plasma homocysteine is elevated in patients lated to systemic vascular dysregulation. Br homocysteine as a risk factor for vascular dis- with exfoliation syndrome. Am J Ophthal-

J Ophthalmol 2004; 88: 662–666. ease. Probable benefits of increasing folic acid mol 2003; 136: 41–46.

2 Kirwan RP, Leonard MO, Murphy M, Clark intakes. JAMA 1995; 274: 1049–1057. 20 Visontai Z, Merisch B, Kollai M, Hollo G: In- AF, O’Brien CJ: Transforming growth fac- 11 Garcia-Tevijano ER, Berasain C, Rodriguez crease of carotid artery stiffness and de- tor-beta-regulated gene transcription and JA, Corrales FJ, Arias R, Martin-Duce A, et crease of baroreflex sensitivity in exfoliation protein expression in human GFAP-negative al: Hyperhomocysteinemia in liver cirrhosis: syndrome and glaucoma. Br J Ophthalmol

lamina cribrosa cells. Glia 2005; 52: 309–324. mechanisms and role in vascular and hep atic 2006; 90: 563–567.

3 Moore P, El-sherbeny A, Roon P, Schoenlein fibrosis. Hypertension 2001; 38: 1217–1221. 21 Vine AK, Stader J, Branham K, Musch DC, PV, Ganapathy V, Smith SB: Apoptotic cell 12 Miller A, Mujumdar V, Palmer L, Bower JD, Swaroop A: Biomarkers of cardiovascular death in the mouse retinal ganglion cell layer Tyagi SC: Reversal of endocardial endothe- disease as risk factors for age-related macu-

is induced in vivo by the excitatory amino lial dysfunction by folic acid in homocyste- lar degeneration. Ophthalmology 2005; 112:

acid homocysteine. Exp Eye Res 2001; 73: 45– inemic hypertensive rats. Am J Hypertens 2076–2080.

57. 2002; 15: 157–163. 22 Pianka P, Almog Y, Man O, Goldstein M, 4 Wong TT, Mead AL, Khaw PT: Prolonged 13 Dalal S, Parkin SM, Homer-Vanniasinkam Sela BA, Loewenstein A: Hyperhomocys- antiscarring effects of ilomastat and MMC S, Nicolaou A: Effect of homocysteine on cy- tinemia in patients with nonarteritic anteri- after experimental glaucoma filtration sur- tokine production by human endothelial or ischemic , central retinal

gery. Invest Ophthalmol Vis Sci 2005; 46: cells and monocytes. Ann Clin Biochem artery occlusion, and central retinal vein

2018–2022. 2003; 40: 534–541. occlusion. Ophthalmology 2000; 107: 1588– 5 Kottler UB, Junemann AG, Aigner T, Zenkel 14 Bleich S, Junemann A, Von Ahsen N, Lausen 1592. M, Rummelt C, Schlotzer-Schrehardt U: B, Ritter K, Beck G, et al: Homocysteine and 23 Cahill M, Karabatzaki M, Meleady R, Ref- Comparative effects of TGF-beta 1 and TGF- risk of open-angle glaucoma. J Neural sum H, Ueland P, Shields D, et al: Raised

beta 2 on extracellular matrix production, Transm 2002; 109: 1499–1504. plasma homocysteine as a risk factor for ret- proliferation, migration, and collagen con- 15 Roedl JB, Bleich S, Reulbach U, Rejdak inal vascular occlusive disease. Br J Ophthal-

traction of human Tenon’s capsule fibro- R, Naumann GO, Kruse FE, et al: Vita- mol 2000; 84: 154–157. blasts in pseudoexfoliation and primary min deficiency and hyperhomocysteinemia 24 Altintas O, Maral H, Yuksel N, Karabas VL,

open-angle glaucoma. Exp Eye Res 2005; 80: in pseudoexfoliation glaucoma. J Neural Dillioglugil MO, Caglar Y: Homocysteine

121–134. Transm 2007; 114: 571–575. and nitric oxide levels in plasma of patients 6 Kuppens EV, van Best JA, Sterk CC, de Kei- 16 Bleich S, Roedl J, Von Ahsen N, Schlotzer- with pseudoexfoliation syndrome, pseudo- zer RJ: Decreased basal tear turnover in pa- Schrehardt U, Reulbach U, Beck G, et al: exfoliation glaucoma, and primary open-an- tients with untreated primary open-angle Elevated homocysteine levels in aqueous gle glaucoma. Graefes Arch Clin Exp Oph-

glaucoma. Am J Ophthalmol 1995;120: 41– humor of patients with pseudoexfoliation thalmol 2005; 243: 677–683.

46. glaucoma. Am J Ophthalmol 2004; 138: 162– 25 Wang G, Medeiros FA, Barshop BA, Weinreb 7 Luo L, Li DQ, Doshi A, Farley W, Corrales 164. RN: Total plasma homocysteine and prima- RM, Pflugfelder SC: Experimental dry eye 17 Roedl JB, Bleich S, Reulbach U, von Ahsen N, ry open-angle glaucoma. Am J Ophthalmol

stimulates production of inflammatory cy- Schlotzer-Schrehardt U, Rejdak R, et al: Ho- 2004; 137: 401–406. tokines and MMP-9 and activates MAPK mocysteine levels in aqueous humor and 26 Cumurcu T, Sahin S, Aydin E: Serum homo- signaling pathways on the ocular surface. plasma of patients with primary open-angle cysteine, vitamin B12 and folic acid levels in

Invest Ophthalmol Vis Sci 2004; 45: 4293– glaucoma J Neural Transm 2007;114: 445– different types of glaucoma. BMC Ophthal-

4301. 450. mol 2006; 6: 6. http://www.biomedcentral. 8 Mujumdar VS, Aru GM, Tyagi SC: Induction 18 Junemann AG, von Ahsen N, Reulbach U, com/1471–2415/6/6. of oxidative stress by homocyst(e)ine im- Roedl J, Bonsch D, Kornhuber J, et al: C677T 27 Mossbock G, Weger M, Faschinger C, Stein- pairs endothelial function. J Cell Biochem variant in the methylentetrahydrofolate re- brugger I, Temmel W, Schmut O, et al: Meth-

2001; 82: 491–500. ductase gene is a genetic risk factor for pri- ylenetetrahydrofolatereductase (MTHFR) 9 Graham IM, Daly LE, Refsum HM, Robin- mary open-angle glaucoma. Am J Ophthal- 677C1 T polymorphism and open angle

son K, Brattstrom LE, Ueland PM, et al: Plas- mol 2005; 139: 721–723. glaucoma. Mol Vis 2006; 12: 356–359. ma homocysteine as a risk factor for vascular disease. The European Concerted Action

Project. JAMA 1997; 277: 1775–1781.

Homocysteine in Primary Open-Angle Ophthalmic Res 2008;40:249–256 255 Glaucoma 28 Fujita M, Igarashi T, Kurai T, Sakane M, 31 Jacques PF, Bostom AG, Wilson PW, Rich S, 33 Choy CK, Cho P, Chung WY, Benzie IF: Wa- Yshino S, Takahashi H: Correlation between Rosenberg IH, Selhub J: Determinants of ter-soluble antioxidants in human tears: ef- dry eye and rheumatoid arthritis activity. plasma total homocysteine concentration in fect of the collection method. Invest Oph-

Am J Ophthalmol 2005; 140: 808–813. the Framingham Offspring cohort. Am J thalmol Vis Sci 2001; 42: 3130–3134.

29 Gulati A, Sullivan R, Buring JE, Sullivan DA, Clin Nutr 2001; 73: 613–621. 34 Callender M, Morrison PE: A quantitative Dana R, Schaumberg DA: Validation and re- 32 Detsky AS, McLaughlin JR, Baker JP, John- study of human tear proteins before and af- peatability of a short questionnaire for dry ston N, Whittaker S, Mendelson RA, et ter adaption to non-flexible contact lenses.

eye syndrome. Am J Ophthalmol 2006; 142: al: What is subjective global assessment of Am J Optom Physiol Opt 1974; 51: 939–945. 125–131. nutritional status? JPEN J Parenter Enteral 35 Van Setten GB, Stephens R, Tervo T, Salonen

30 Schiffman RM, Christianson MD, Jacobsen Nutr 1987; 11: 8–13. EM, Tarkkanen A, Vaheri A: Effects of the G, Hirsch JD, Reis BL: Reliability and valid- Schirmer test on the fibrinolytic system in

ity of the Ocular Surface Disease Index. Arch the tear fluid. Exp Eye Res 1990; 50: 135–141.

Ophthalmol 2000; 118: 615–621.

256 Ophthalmic Res 2008;40:249–256 Roedl et al.