Examples of endothelial morphometricassessment by grading scales. A, Mean product score of 11.2: score 4 (<5%dead cells) × score 2.8 (completely normal morphologic characteristicsaccording to 4 of 5 observers). B, Mean product score of 6.4: score 4 (<5%dead cells) × score 1.6 (abnormal to dubious morphologic characteristics,poor swelling [star], and Descemet fold [arrowheads]). C, Mean product scoreof 6.2: score 2.6 (5%-25% dead cells [white arrowheads], over Descemet foldsand diffusely) × score 2.4 (Descemet fold [black arrowheads], relativelynormal morphologic characteristics in areas without dead cells). D, Mean productscore of 2.6: score 1.6 (±25% dead cells [white arrowheads], diffuselyand over Descemet fold) × score 1.6 (abnormal morphologic features:poor swelling [stars] and Descemet fold [black arrowheads]).
Cytotoxicity indices of humancorneal fibroblast cultures after exposure to trypan blue. A, Eagle modifiedminimum essential medium vehicle. B, Phosphate-buffered saline vehicle. Thegray bars indicate significantly toxic cytotoxicity indices (cutoff point,20%). Bars with black tops indicate indices of less than 0%.
Example of a sigmoid concentration-responsecurve of human corneal fibroblasts exposed to trypan blue (in this case, 18-hourexposure in Eagle modified minimum essential medium [EMEM] vehicle [dashedcurve] and phosphate-buffered saline [PBS] vehicle [solid curve]). The x-axisuses a logarithmic scale.
van Dooren BTH, Beekhuis WH, Pels E. Biocompatibility of Trypan Blue With Human Corneal Cells. Arch Ophthalmol. 2004;122(5):736-742. doi:10.1001/archopht.122.5.736
Copyright 2004 American Medical Association. All Rights Reserved.Applicable FARS/DFARS Restrictions Apply to Government Use.2004
To quantify the toxicity of trypan blue on human corneal cells accordingto exposure time and concentration.
Three in vitro experiments were performed. (1) We exposed cultured humancorneal fibroblasts to trypan blue (0.0001% to 0.1%) in Eagle modified minimumessential medium (EMEM) or phosphate-buffered saline (PBS) for 15 minutesto 24 hours. Cytotoxicity was evaluated by Mosmann's colorimetric 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MMT) assay. (2) We exposed human corneas in EMEM for24 hours to trypan blue (0.001% to 0.1%). Fellow donor corneas served as controls.Endothelial survival was evaluated morphologically and by cell density assessment.(3) We morphologically compared the endothelial viability of human donor corneasafter exposure to 0.1% trypan blue for 5 to 30 minutes with control corneas.
In experiment 1, trypan blue in EMEM was not significantly toxic atconcentrations of 0.005% or lower. Higher concentrations were toxic only afterexposure to trypan blue for at least 6 hours. In PBS, significant toxicitywas found after exposure to 0.1% trypan blue for 30 minutes or longer. Lowerconcentrations were toxic after longer exposures. In experiment 2, exposureto 0.01% and 0.1% trypan blue for 24 hours resulted in significant loss incell density. At lower concentrations, the endothelium was affected only morphologically.In experiment 3, endothelial morphology changed in control corneas and afterexposure to 0.1% trypan blue for as little as 5 minutes. After 30-minute exposure,morphologic deterioration was more pronounced.
Trypan blue was toxic in vitro to corneal endothelium and corneal fibroblastsat higher concentrations and notably longer exposure times. Toxicity was lessin EMEM than in PBS.
At commonly used concentrations, both during cataract surgery and inthe cornea bank, trypan blue is safe for corneal cells. At higher concentrationsor longer exposures, however, caution is warranted.
Trypan blue vital staining for the evaluation of the endothelium ofdonor corneas was first described by Stocker et al.1 Trypanblue stains the nuclei of severely damaged and dead endothelial cells of donorcorneas, as well as areas of Descemet membrane denuded of endothelial cells.It does not stain viable endothelial cells with an intact cell membrane.1- 4 Sperlinget al5- 7 introducedtrypan blue vital staining in combination with induced dilation of intercellularspaces (with 0.45% and 0.9% sodium chloride, or with 1.8% sucrose) for visualizationof endothelial cell borders, thus allowing light-microscopic assessment ofboth endothelial damage and endothelial cell density (ECD) on potentiallytransplantable donor corneas. Since 1982, this technique has been used bythe Cornea Bank of the Netherlands Ophthalmic Research Institute (NORI) forthe evaluation of the endothelium of human donor corneas after storage byorgan culture preservation.8,9 Variationsof this technique, all using trypan blue, are applied in many cornea banksin Europe.
Norn10- 12 wasthe first to use trypan blue intraoperatively, during intracapsular cataractsurgery, to evaluate endothelial status. Recently, intraoperative applicationof the dye has gained new interest: trypan blue anterior lens capsule stainingaims at improving visualization of the capsulorrhexis during phacoemulsificationin patients with absent red fundus reflex, eg, in mature cataracts.13
To expand the existing body of knowledge on the toxicity of trypan blueon human corneal cells, concentration- and exposure time–related effectsof trypan blue were studied in 3 in vitro experiments on human donor corneasand human corneal fibroblasts.
Trypan blue (Gurr; BDH Chemicals Ltd, Poole, England) was dissolvedin sterile phosphate-buffered saline (PBS) (Azua Pharmacy, Amsterdam, theNetherlands) unless otherwise stated. The Cornea Bank NORI has used a 1.2%solution from this stock for routine purposes since 1998. The photometricextinction value of this solution is 0.227 at 560 nm. This value is the sameas for the 0.3% trypan blue solution, obtained from another stock, and describedand applied in previous years.8,14,15
Corneal fibroblasts were obtained from normal human donor eyes as describedpreviously.16,17 In short, cornealbuttons were excised and fragmented. Three to 4 corneal pieces were implantedper well of a 6-well plate (Nunc, Roskilde, Denmark) and incubated with 1mL of Eagle modified minimum essential medium (EMEM) (ICN Biomedicals Inc,Costa Mesa, Calif) supplemented with 10% fetal bovine serum (FBS) (Gibco LifeTechnologies Ltd, Paisley, Scotland), 100-U/mL penicillin (Gist Brocades,Leiderdorp, the Netherlands), and 50-µg/mL streptomycin (Biochemie GmbH,Vienna, Austria) at 37°C. When the pieces adhered firmly to the bottomof the well by the outgrowing cells, 5 mL of medium per well was added. Mediumwas renewed twice a week. Outgrowing fibroblasts were subcultured after 25to 28 days and fibroblasts of the second passage were stored in liquid nitrogen.Cells were defrosted, subcultured for 1 passage, and used for experiments.At confluency, the cells were removed from the culture flasks by incubationwith a trypsin-EDTA solution, collected by centrifugation (200g), resuspended in medium, and plated out at 5 × 104 cellsper well in 96-well flat-bottomed tissue culture plates (Nunc). The cellswere cultured at 37°C in a humidified atmosphere with 5% carbon dioxidebefore exposure to freshly prepared sterile trypan blue solutions, varyingin concentration from 0.0001% to 0.1%, in a vehicle of EMEM (half of the wellsper tray) or sterile PBS. Wells were filled with 100 µL of study preparations,either with the trypan blue solutions as described or vehicle (EMEM or PBS)without added trypan blue as control.
Cytotoxicity was determined with Mosmann's colorimetric 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay directly after exposure to trypan blue.18 The MTT assay was performed as described previously19,20; it basically measures the capacityof mitochondrial enzymes to transform 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MMT) (Sigma-Aldrich Corp, St Louis, Mo) into a formazanproduct and thus reflects the mitochondrial activity of corneal fibroblasts.After exposure of the cells to the different concentrations of trypan blueduring exposure times ranging from 15 minutes to 24 hours, the cells werecarefully washed 3 times with PBS and 100 µL of a culture medium with2% FBS was added per well. Then 10 µL of a freshly prepared MMT solution(5-mg/mL MMT in PBS) was added to all wells and plates were incubated at 37°Cfor 4 hours. The MMT-formazan reaction product was solubilized by the additionof 200 µL of acid isopropanol (0.04N hydrochloric acid in propanol-2)to the incubation medium without washing steps and incubated overnight atroom temperature in the dark. The optical density of the solution was assessedat 560 nm in a spectrophotometer (IEMS Reader MF; Labsystems, Helsinki, Finland),taking medium with MTT and isopropanol as the blank. Each experiment was performedin 4-fold, and the experiments were repeated 3 times. Irreversibility of theeffect of exposure to trypan blue was confirmed by repeating the MMT assay2 to 3 days and 7 to 8 days after exposure to trypan blue.
The cytotoxicityindex (CI) of each culture was calculated by the following formula:
CI= [(ODcontrol–ODtest)/ODcontrol] ×100%,
where ODcontrol is the optical density in the controlwells (cell incubated with vehicle only, PBS or EMEM) and Odtest isthe optical density in the wells with cells incubated with trypan blue inthe vehicles PBS and EMEM.
A CI of higher than 20% was considered to be significant, thus compensatingfor the standard deviation of mean CI from the 3 experiments.19,20
Paired corneas from human donors judged unsuitable for transplantationfor medical reasons or corneal scars were used. One randomly chosen corneafrom each pair was used as a control and stored at 31°C in EMEM with 2%FBS and antibiotics, whereas the experimental cornea was preserved at thesame temperature in the same medium, to which trypan blue was added. Concentrationsof trypan blue that resulted were as follows: 0.1% (2 pairs), 0.01% (2 pairs),0.005% (5 pairs), and 0.001% (5 pairs). After exposure for 24 hours, all corneaswere rinsed and preserved in fresh EMEM at 31°C for another 21 days. At9 to 10 and 20 to 21 days, corneal endothelium was evaluated by the techniciansof the cornea bank. Storage of donor corneas in EMEM, as well as endothelialevaluation, was performed under sterile conditions using standard eye banktechniques, as described previously.8,14,15 Endotheliaof experimental and control corneas were compared on ECD and morphologic status.Endothelial morphology was graded normal (score of 3) when a normal, "stringof pearls"–like swelling pattern of the intercellular space indicativeof a well-functioning cell membrane was present; dubious (score of 2) wasassigned to endothelia with irregular or locally poor swelling of the intercellularspace, and with the presence of many vacuoles within the endothelial cells;and abnormal morphology (score of 1) was used for endothelia with roundingup of the cells, indicative of loss of intercellular contacts, and with generallypoor swelling of the intercellular space (endothelia were called dead, witha score of 0, when no recognizable endothelial cells could be discerned).14,15
Statistical evaluation of ECD was performed by means of a mixed-modelanalysis of variance with ECD as the dependent variable, adjusted for itsbaseline measurements and time of investigation as covariate in the model.There was a within-subject factor (trypan blue vs control) and a between-subjectsfactor (concentration) in the model. Interest was in the interaction betweenthese latter 2 factors. No structure was assumed for the within-subject correlations.Endothelial morphology scores were evaluated by means of a nonparametric Wilcoxonsigned rank test. Experimental corneas, all taken together, were tested againstall control corneas for both time points. Testing by concentration group wasconsidered not to be meaningful because of small group size.
Human donor corneas were used that were judged unsuitable for transplantationbecause of corneal scars, low ECD (between 1800 and 2300 cells/mm2),or signs of abnormal morphologic characteristics such as significant polymegethism.14,15 These corneas were briefly (1-2 days)stored at the Cornea Bank NORI under sterile conditions at 31°C in EMEMwith 2% FBS and supplemented with 5% dextran before shipment to the RotterdamEye Hospital, Rotterdam, the Netherlands, for the experiment.
Both at the beginning and at the end of the experiment, the endotheliumof all corneas was evaluated by 0.1% trypan blue vital staining and provokedintercellular swelling with 1.8% sucrose solution (hospital pharmacy of theRotterdam Eye Hospital), and photographed with an inverted light microscope.Corneas with endothelia that showed severe degenerative changes (ie, trypanblue–stained nuclei, vacuoles in the cells, poor swelling of the intercellularspace) before the experiment were excluded from the experiment. The remainingcorneas were randomly allotted to one of the treatment groups or the controlgroup.
The control corneas (n = 8) were merely incubated for 2 hours in EMEMwith 2% FBS and 5% dextran at room temperature. In 3 of the experimental groups,0.1% trypan blue solution was applied to the endothelium of the corneas directlyafter the preexperimental evaluation for 30 (n = 7), 10 (n = 9), and 5 (n= 7) minutes. A fourth experimental group (n = 6) was exposed to 0.01% trypanblue for 10 minutes. Then, after trypan blue was thoroughly rinsed off withbalanced saline solution, all experimental corneas were incubated in EMEMfor 2 hours with 2% FBS and 5% dextran at room temperature.
The preexperiment and postexperiment photomicrographs of the endotheliumof control and experimental corneas were mixed and subjected to a masked assessmentby 5 laboratory technicians of the Cornea Bank NORI. They used a 4-tieredgrading scale to estimate the percentage of dead endothelial cells (4, <5%dead cells; 3, <5% dead cells, but dead cells mainly over folds in Descemetmembrane; 2, 5%-25% dead cells, either diffusely or over folds, or both, and1, >25% dead cells). The same morphologic grading scale as in the second,"paired" experiment was used to assess morphologic features of areas withoutdead cells (3, normal; 2, dubious; and 1, abnormal morphologic characteristics).For each photomicrograph and for each of the observers, a product score rangingbetween 1 and 12 was calculated from the scores on the 2 scales. Interobservercorrelation was evaluated by calculating Cronbach α. By averaging theproduct scores of the 5 observers, an endothelial morphology assessment scorewas obained for each cornea before and after the experiment. The product scorewas considered to be the best means to reflect the viability of the endothelium,irrespective of cell density or variation in cell size. Figure 1 demonstrates this grading system by showing 4 photomicrographsof donor cornea endothelia used in this experiment with their resulting meanendothelial assessment scores.
The differences between preexperiment and postexperiment endothelialmorphology assessment scores were calculated and expressed as change scores.With these morphologic change scores, a linear regression model was constructed,and contributory effects of trypan blue exposure time were evaluated. Theeffects on the endothelium of the performance of the experiment per se werealso investigated.
The relationship between the mean CI, trypan blue concentration, andexposure time is shown in the 3-dimensional graphs in Figure 2A and B for EMEM and PBS, respectively. Significant CIs(>20%) are represented by the gray bars in the graphs. Figure 2A shows that in EMEM, trypan blue concentrations of 0.01%and higher resulted in significant toxicity after exposure for 6 hours orlonger. At 24 hours' exposure, a trypan blue concentration of 0.005% showeda threshold significant CI. Lower concentrations did not cause significanttoxicity with exposures up to 24 hours. The graph in Figure 2B shows that in PBS significant toxicity was observed atexposures to 0.1% trypan blue of 30 minutes or longer. The exposure timesneeded to reach significant CIs increased with decreasing trypan blue concentrations.With 0.0001% trypan blue in PBS, no significant CI was observed, regardlessof exposure times.
Dose-response relationships for trypan blue in the 2 vehicles, EMEMand PBS, were plotted for each of the different exposure times. Response wasexpressed as the mean CI obtained in the 3 (repeated) experiments. Figure 3 shows a representative dose-responsecurve, in this example after 18 hours' exposure to trypan blue in EMEM. Usinga sigmoidal fit program (Origin, version 4.0; Microcal Software Inc, Northampton,Mass), the maximal toxicity (CI) levels and LC50s (50% lethal concentration,ie, concentration resulting in half-maximal response) were calculated forall exposure times and for both of the vehicles. These data are summarizedin Table 1. The LC50 appearedto be independent of exposure time and vehicle and was found at a concentrationof 0.02% to 0.04%. In contrast, the calculated maximal toxicity levels werestongly dependent on exposure time, and higher calculated maximal toxicitieswere found in PBS than in EMEM.
At days 2 to 3 and days 7 to 8 after exposure, the results (not shown)were essentially the same as the results directly after exposure, indicatingthat the observed damage to the fibroblasts was irreversible.
The results after 9 to 10 days and 20 to 21 days of preservation ofthe donor cornea pairs were expressed as mean ECDs, endothelial morphologyscores, and percentages of ECD loss (Table2). All corneas exposed to 0.1% or 0.01% trypan blue suffered totalendothelial cell loss at 20 to 21 days of preservation; the endothelium ofone cornea exposed to 0.1% trypan blue had already been destroyed at day 9.
Analysis of variance outcomes showed a significant interaction of thewithin-subject factor (trypan blue vs control) and the between-subject factor(concentration) (P<.001), as well as separatesignificant main effects of the factor "concentration" (P<.001) and the factor "trypan blue vs control" (P<.001). These results indicate a significant concentration-dependenteffect of trypan blue on ECD when experimental corneas were compared withtheir fellow controls. Estimated effects by this model of the interaction(concentration–trypan blue vs control) on ECD loss, with 95% confidencelimits, were 60 cells/mm2 for trypan blue concentration of 0.001%(P = .57), 93 cells/mm2 for 0.005% trypanblue (P = .38), 2078 cells/mm2 for 0.01%trypan blue (P<.001), and 1894 cells/mm2 for 0.1% trypan blue (P<.001) (Table 2).
On both observation points (days 9-10 and days 20-21), the Wilcoxonsigned rank test showed a significant effect of trypan blue (all concentrationstogether) on endothelial cell morphologic characteristics (P = .03 and P = .02, respectively) as comparedwith the endothelial cell morphology scores of all control corneas.
Endothelial morphology assesment and change scores are summarized in Table 3. Cronbach α, a coefficientfor interobserver reliability, was 0.942; this confirms strongly the validityof the grading scales.
Morphologic change scores were normally distributed. Unpaired, 2-tailed t test showed no significant difference (P = .33) between morphologic change scores in the corneas in the controlgroup (n = 8) and the corneas in all the other groups together (n = 29), indicatingan overall detrimental effect on the endothelium of manipulation of the donorcorneas by merely performing this experiment. A linear regression model wasconstructed of the morphologic change values for the 0.1% trypan blue concentration.The overall contribution of exposure time was not significant (P = .18); however, comparison of the different exposure times showedthe following: controls vs 5 minutes, P = .88; 5minutes vs 10 minutes, P = .92; controls vs 30 minutes, P = .08; 5 minutes vs 30 minutes, P =.07; and 10 minutes vs 30 minutes, P = .06. Thisis an indication that exposure to 0.1% trypan blue for 30 minutes may causeadditional deterioration of endothelial morphologic features.
The average morphologic change score of the group corneas exposed to0.01% trypan blue appears to be lower than those of the groups exposed to0.1% trypan blue (Table 3).
The results of our 3 experiments, in which different methods were appliedto assess toxic effects of trypan blue for corneal cells, endothelial cellsin particular, demonstrated the following effects: (1) a concentration-relatedeffect of trypan blue on corneal cell viablity; (2) an effect of exposuretime to trypan blue on cell survival; (3) an effect of the vehicle on thedegree of toxicity exhibited by trypan blue; and (4) damaging effects of (experimental)manipulation of cells or endothelia.
Each of the methods used in the 3 different experiments has its advantagesand disadvantages. The first experiment assessing the cytotoxicity of trypanblue for corneal fibroblasts used an objective, quantitative evaluation methodand homogeneous material of good quality, and had closely controlled experimentalcircumstances. However, no endothelial cells but human fibroblast cell cultureswere used, and therefore this was not a true assessment of the toxicity oftrypan blue on human corneal endothelium in situ. The second, paired experimentmet this latter criterion. Furthermore, its design, using organ culture preservationas a controlled condition "stress" test for endothelial cell survival on pairedexperimental vs control endothelia, provided for a powerful quantitative assessmentof long-term endothelial cell viability after exposure to trypan blue, obviatingthe need for large amounts of tissue. On the downside, relatively subjectiveassessment methods were needed to obtain quantitative measures, a dose-responserelationship was much more difficult to evaluate, and effects of exposuretime were not examined. In the third experiment, several short exposure timeswere studied, and the design, application of trypan blue to donor cornea endothelium,reflected the actual situation in eye banks rather well. However, statisticalevaluation suffered from lack of power due to relatively small numbers ofcorneas per treatment group, because of restricted supply of suitable donormaterial. Second, the assessment of endothelial morphometrics was subjective;however, the use of 5 independent masked experienced observers and standardizedgrading scales that provided for a very high interobserver reproducibilityboth improved on the reproducibility of this semiquantitative assessment.Third, the surprising outcome with exposure for 10 minutes to 0.01% trypanblue (Table 3) indicates a limitationof the use of morphometric measures on viable endothelium. When the osmolarityof the solutions tested differs, eg, by the use of a different concentrationof trypan blue in PBS, as was the case in this experiment, the endotheliummay become temporarily edematous and more vulnerable to a subsequent swellingof the intercellular space necessary to visualize living endothelial cellborders. Therefore, concentration-dependent effects were not further testedin this experiment. Finally, preexperimental status of the endothelia wasrather heterogeneous because of the selection criteria and preexperiment conditions.This was compensated for by carefully evaluating preexperiment morphometricstatus and discarding morphologically clearly abnormal endothelia before theexperiment started. This preexperiment evaluation appeared to have succeeded,for Table 3 shows that the preexperimentendothelial assessment scores were quite good, and similar in all experimentaland control groups.
Although each of the experiments may have had its own flaws, each showedconsistent results that furthermore were complementary to and in concordancewith those from the other experiments. This study demonstrated a consistentrelationship between concentration, exposure time, vehicle of trypan blue,and/or application circumstances. Under favorable conditions (the second experiment,and EMEM in the first experiment), 24 hours of exposure to a trypan blue concentrationof 0.005% was found to be a threshold at and below which there was no significanttoxicity. Under unfavorable conditions (PBS in the first experiment, and therepeated manipulation in the third experiment), the toxicity threshold wasfound at an exposure to 0.1% trypan blue for 30 minutes; this threshold shiftedtoward longer exposure times as concentrations decreased and vice versa.
A recent experimental study21 confirmsconcentration-related toxicity and the existence of a threshold of toxicityrelated to trypan blue application. Interest had arisen in the applicationof trypan blue as a surgical tool to facilitate visualization in epiretinalmembrane surgery. Trypan blue application in a 0.06% concentration in thevitreous cavity of rabbit eyes for up to 4 weeks did not cause any recognizabledamage to the rabbit retina. In contrast, the application of 0.2% trypan bluecaused considerable damage to retinal cells and architecture in areas thatwere more exposed to the dye.21 It is, however,not feasible to compare in more detail the results of this study with ourown findings on endothelial cells because of the different nature of the cellsof interest, the different circumstances under which the exposure took place(in vivo vs in vitro), the far longer exposure times used, and the differenttechniques applied for the determination of toxicity.21
The results in our study seem to be in line with the literature andwith experience with the dye in eye banks. Sperling4 andStocker et al22 stained the endothelium ofhuman donor corneas briefly (1 and 1.5 minutes, respectively) with 0.3% andwith 0.25% to 0.5% trypan blue solutions, respectively. Both authors foundno indications for endothelial cell loss caused by the staining.4,22 Formany years, the Cornea Bank NORI applied 0.3% trypan blue to the endotheliumfor 30 to 60 seconds for routine donor evaluation.14,15 In1998, colorimetric investigations of a new stock of trypan blue disclosedhigher extinction values of the 0.3% solutions than before; the concentrationof trypan blue solution for donor cornea evaluation was then changed to 0.12%to use a solution with the same extinction value as before. For almost 20years and more than 14 000 transplanted corneas, adverse effects to theendothelium have not been observed with application of 0.3% or 0.12% trypanblue solutions (E.P., unpublished results, 2000). Our present experimentalfindings may be considered to be congruous with our experience with trypanblue in the Cornea Bank NORI and an extrapolation of the findings in the olderstudies. Although the reported concentrations of trypan blue were often higherthan those used in our experiments, endothelia were exposed for shorter times,which probably prevented possible toxic effects from taking place.
Observed toxicity limits in this study allow the intraoperative applicationof trypan blue. In lens capsule staining with trypan blue during cataractsurgery, a concentration of 0.06% is used to obtain adequate visualizationof the lens capsule in the absence of a red fundus reflex. The dye is appliedto the anterior lens capsule under an air bubble in the anterior chamber,which prevents dilution of the trypan blue by aqueous before staining thecapsule and also impairs to some extent direct contact of the dye with theendothelium. Immediately after staining, the excess dye is washed out of theanterior chamber by copious irrigation, after which a viscoelastic substanceis injected. The dye on the anterior lens capsule is partly removed with theanterior lens capsule after the capsulorrhexis is completed. The remainingdye is partly washed out with irrigation during the operative procedure, andremaining traces are thought to leave the anterior chamber by the aqueousroute. The dye cannot be observed anymore on the first postoperative day.13 Thus, an initial concentration of 0.06% trypan bluemay partly contact the endothelium for a few seconds. Thereafter, only dilutedand steadily decreasing concentrations of trypan blue are thought to be presentwithin the anterior chamber because of a continous washout effect throughaqueous flow and resorption. The results of our in vitro study support theview that concentrations of trypan blue in the anterior chamber after anteriorlens capsule staining with trypan blue in cataract surgery do not reach atoxic level. Confirmation of safety for the corneal endothelium of trypanblue capsule staining was published recently in a prospective, randomizedclinical trial on endothelial cell loss. At 1 year postoperatively, no deleteriouseffects could be observed, either on corneal ECD or on endothelial morphometricmeasures.23
Corresponding author: Bart T. H. van Dooren, MD, Rotterdam Eye Hospital,PO Box 70030, 3000 LM Rotterdam, the Netherlands (e-mail: firstname.lastname@example.org;email@example.com).
Submitted for publication February 3, 2003; final revision receivedOctober 22, 2003; accepted November 21, 2003.
We thank the following persons: Jikke Kingma, former laboratory technicianat the Cornea Bank NORI, Amsterdam, the Netherlands, for assistance in performingthe first and second experiments; all laboratory technicians of the CorneaBank NORI, for assistance in assessing the photographs in the third experiment;Paul G. H. Mulder, PhD, Department of Epidemiology and Biostatistics, ErasmusMedical Centre, Rotterdam, the Netherlands, for statistical consultation inthe second and third experiments; and Gerrit R. J. Melles, MD, PhD, the NetherlandsInstitute for Innovative Ocular Surgery, Rotterdam, for valuable advice.