Ability of New Vital Dyes to Stain Intraocular Membranes and Tissues in Ocular Surgery EDUARDO B. RODRIGUES, FERNANDO M. PENHA, ELAINE DE PAULA FIOD COSTA, MAURICIO MAIA, EDUARDO DIB, MILTON MORAES, JR, CARSTEN H. MEYER, OCTAVIANO MAGALHAES, JR, GUSTAVO BARRETO MELO, VINICIUS STEFANO, ANA BEATRIZ DIAS, AND MICHEL EID FARAH ● PURPOSE: To evaluate the ability of novel dyes to stain lens capsule (LC), internal limiting membrane (ILM), epiretinal membrane (ERM), and vitreous. ● DESIGN: Experimental study in animal and human donor eyes. ● METHODS: Thirteen dyes, methyl violet, crystal violet, eosin Y, sudan black B, methylene blue, toluidine blue, light green, indigo carmine, fast green, congo red, evans blue, brilliant blue, and bromophenol blue, were injected onto the LC and ILM of enucleated porcine eyes. The vitreous was stained with 2 mL of dyes for 1 minute. Six dyes (indigo carmine, evans blue, fast green, light green, bromophenol blue, and brilliant blue) were selected for experiments in human donor eyes and freshly removed ERM. ● RESULTS: In the porcine eyes, ILM staining with methylene blue, toluidine blue, indigo carmine, evans blue, bromophenol blue, and fast green was moderate, and methyl violet, crystal violet, brilliant blue, or sudan black resulted in strong staining. Methyl violet, crystal violet, sudan black, toluidine blue, and methylene blue caused histologic damage in porcine retinas. Vitreous examination revealed moderate staining with congo red, crystal violet, fast green, eosin Y, methylene blue, toluidine blue, brilliant blue, bromophenol blue, and methyl violet and strong staining with light green and evans blue. ERMs showed strong staining with 0.5% evans blue and moderate staining with 0.5% light green, fast green, brilliant blue, and bromophenol blue. Evaluation of donor eyes disclosed moderate staining with evans blue, light green, and bromophenol blue and strong staining with 0.5% brilliant blue. Moderate or strong staining of the vitreous occurred with most dyes. LC evaluation showed moderate staining with 0.5% evans blue, fast green, and brilliant blue, whereas 0.5% light green produced strong LC staining. ● CONCLUSIONS: Brilliant blue shows the best ILM staining, whereas bromophenol blue, evans blue, and
Accepted for publication Aug 18, 2009. From the Vision Institute (IPEPO), Department of Ophthalmology, Federal University of São Paulo, São Paulo, Brazil (E.B.R., F.M.P., E.d.P.F.C., M.M., E.D., M.M.Jr., O.M., G.B.M., V.S., A.B.D., M.E.F.); and the Department of Ophthalmology, University of Bonn, Bonn, Germany (C.H.M.). Inquiries to Michel Eid Farah, R. Botucatu 820, 04023-062 São Paulo SP, Brazil; e-mail: [email protected]
light green also stain ILM. Most dyes bind well to LC, vitreous, and ERM. (Am J Ophthalmol 2010;149: 265–277. © 2010 by Elsevier Inc. All rights reserved.)
NTRAOPERATIVE APPLICATION OF VITAL DYES FOR THE
visualization of intraocular membranes and tissues has facilitated surgical techniques and outcomes in recent years.1 For vitreoretinal surgery, indocyanine green (ICG) dye initially was introduced for staining the internal limiting membrane (ILM).2 This was followed by many publications reporting signs of retinal toxicity after intravitreous ICG injection.3– 6 In cataract surgery, trypan blue has provided since the 1990s a better visualization of the anterior lens capsule (LC) when the red reflex is not possible.1 Following ICG and trypan blue, other dyes such as brilliant blue and patent blue were presented as newer alternative vital dyes for ocular surgery.7,8 However, concerns also were raised about the safety of brilliant blue, trypan blue, and patent blue, and their selective affinity for some intraocular membranes such as epiretinal membranes (ERMs), LC, or ILM, but not for all of them.9 –11 For this study, a total of 13 vital dyes were selected: light green, fast green, methyl violet, crystal violet, congo red, eosin Y, sudan black B, evans blue, brilliant blue, bromophenol blue, methylene blue, toluidine blue, and indigo carmine. In addition, ICG, trypan blue, and patent blue were included in our in vivo examinations for comparison. The goal of this study was to investigate the staining ability of these dyes regarding the retinal ILM, vitreous, ERM, and LC in a series of systematic experiments in freshly enucleated porcine eyes, freshly removed ERM, and enucleated human donor eyes.
METHODS ● SELECTION OF DYES AND PREPARATION:
A total of 50 mg of dye powder of light green, fast green, bromophenol blue, brilliant blue, patent blue, methyl violet, crystal violet, congo red, eosin Y, sudan black, evans blue, methylene blue, toluidine blue, indigo carmine, ICG, or trypan blue (Merck, Darmstadt, Germany, for all dyes, except for Sigma-Aldrich, Munich, Germany, in the case of bromophenol blue and Ophthalmos Ind, São Paulo, Brazil, for ICG and trypan blue) was weighed with an
ELSEVIER INC. ALL
analytic balance (Mettler-Toledo, Inc, Columbus, Ohio, USA) and dissolved in 10 mL balanced salt solution (BSS plus; Alcon Laboratories, Inc, Fort Worth, Texas, USA) to obtain a concentration of 0.5% for the initial experiments. The mixtures were shaken for 5 minutes and sonicated (Unique Ind, Idaiatuba, Brazil) to obtain a homogeneous solution. For donor eye experiments, the selected dyes light green, ICG, trypan blue, patent blue, fast green, indigo carmine, evans blue, brilliant blue, and bromophenol blue were prepared as 0.5% and 0.05% solutions in balanced salt solution. The choice of agents was based on a rationale to validate the experimental method; thiazine, carbonyl, diazo, aminoarylmethane, and hydroxyxanthene dyes were analyzed to search for an optimal staining agent.12
as no staining (0), faint staining (⫹), moderate staining (⫹⫹), or strong staining (⫹⫹⫹).13 ● HISTOLOGIC EVALUATION OF PORCINE EYES BY
LCs were placed in 10% buffered formaldehyde and 4% buffered glutaraldehyde (Karnowsky fixative) for a minimum of 48 hours. Specimens of LC then were evaluated macroscopically by light microscopy after preparation with routine hematoxylin and eosin staining. Immediately after ILM removal, some specimens of both ILM and peeled retinas of porcine eyes were placed in a mixture of 4% buffered formaldehyde and 4% buffered glutaraldehyde (Karnowsky fixative) for 15 to 20 minutes to achieve rapid and proper fixation of the retina followed by storage in 10% buffered formaldehyde for 48 hours. Gross examinations of the tissues were performed. Semithin sections were cut along a superior–inferior plane, and they then were embedded in paraffin. Next, the retinas were sectioned at a thickness of 6 m, stained using routine hematoxylin and eosin and examined by light microscopy. In screening for the most suitable dye for experiments in donor eyes, each dye was considered to be toxic to the retina if any of the following findings were observed: major signs of disorganization, cell loss, vacuolization, condensation, edema or swelling; necrosis of at least one retinal layer; or fragments, debris, as well as pieces of retinal cell elements within the removed ILMs.
● STAINING OF PORCINE INTERNAL LIMITING MEMBRANE, VITREOUS, AND LENS CAPSULE: All animal experiments were conducted according to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The experiments for ILM, LC, and vitreous staining were conducted on an ex vivo setup. Ninety-six pig eyes were obtained from a slaughterhouse and were prepared within 6 hours after death. The dyes initially were used to stain the LC of pigs. Immediately after eye removal, the eyes were placed on a glass support, and a 360-degree limbal incision was performed for corneal removal, followed by total iris removal. Next, 0.05 mL of 2 concentrations of 0.05% and 0.5% of each of the 16 dyes were placed onto the LC for 1 minute. Afterward, the anterior surface of the lens was irrigated carefully with saline solution. An anterior curvilinear continuous capsulorrhexis was performed within the stained capsule using forceps. For evaluation of ILM and vitreous staining, the remaining eye cups had the anterior segment removed with a 3-mm incision behind the limbus, and eyes with funduscopic red reflex were included in the experiments. The vitreous was gently extracted completely with cotton buds and forceps and was placed in a glass vial for its evaluation. Next, 0.05 mL of 0.5% of the 16 dyes was injected slowly over the retinal surface to avoid dispersion. The dye was left on the retina for 2 minutes and the retinal surface was rinsed twice with 2 mL saline. Each experiment was repeated with 4 samples with each dye. In all eyes, peeling of the ILM was attempted with a bent 27-gauge needle. To evaluate vitreous staining, the vitreous removed from fresh enucleated pig eyes was placed in a 5-mL vial for 1 minute with 2 mL of each of the 13 dyes at a concentration of 0.5%. The vitreous then was removed, rinsed thoroughly, and placed on a white background for staining evaluation under the microscope. Photographs of all experiments were obtained with a Canon camera (Canon PowerShot A2000 IS; Canon Electronics, Saitama, Japan) fitted to the operating microscope. The intensity of ILM, LC, and vitreous staining was graded in a masked fashion
● STAINING OF FRESHLY EXCISED HUMAN EPIRETINAL
This study was approved by local ethical committee and informed consent was obtained from each patient before surgery. Staining pattern of human ERMs was examined after exposure to 0.5% of the 16 dyes included in this study. The vital dyes (0.3 mL) were applied onto freshly surgically removed ERMs obtained from patients with proliferative diabetic retinopathy and proliferative vitreoretinopathy of rhegmatogenous retinal detachment. The staining intensity of ERMs was graded as for other membranes in as masked fashion. The stained ERM tissues also were fixed with 10% buffered formaldehyde followed by hematoxylin and eosin preparation for light microscopy to detect vitreous or ILM remnants and the composition of the tissue.
● STAINING OF LENS CAPSULE, VITREOUS, AND INTERNAL LIMITING MEMBRANE IN HUMAN DONOR EYES:
Human donor eyes were obtained from the Eye Bank of the Federal University of São Paulo after institutional review board approval. Six dyes were chosen for this setup based on their safety and staining profile on the porcine eyes: light green, fast green, brilliant blue, indigo carmine, bromophenol blue, and evans blue. In addition, analysis of staining in these donor eyes was conducted with ICG, trypan blue, and patent blue to compare the currently used dyes with the novel future dyes. Initially, human cadaveric OF
VOL. 149, NO. 2 TABLE 1. Staining Properties of Vital Dyes in Porcine Eyes Ex Vivo Histologic Results at 0.5%
Ex Vivo Staining in Porcince Eyes at 0.5% Biochemical Properties and Principles of Biological Staining
NEW VITAL DYES IN
Lens Capsule Staining
Color on the Retina
Retina Histologic Damage
Methylene blue Toluidine blue Indigo carminea Evans bluea Congo red Light green Sfa Eosin Yb Sudan black B Crystal violet Fast greena Methyl violet Patent blue Indocyanine green Trypan blue Brilliant bluea Bromophenol bluea
Thiazine Thiazine Carbonyl Anionic dis-azo Anionic dis-azo Anionic amino triarylmethane Acidic hydroxyxanthene Anionic dis-azo Cationic amino arylmethane Anionic amino arylmethane Cationic amino arylmethane Anionic amino arylmethane Tricarbocyanine Anionic dis-azo Triarylmethane Triarylmethane
C16H18N3ClS C15H16N3SCl C16H8N2Na2O8S2 C34H24N6Na4O14S4 C32H22N6Na2O6S2 C37H34N2Na2O9S3 C20H8O5Br4 C29H24N4 C25H30N3 C37H34N2Na2O10S3 C24H28ClN3 C27H31N2NaO6S2 C43H47N2NaO6S2 C34H24N6Na4O14S4 C47H48N3S2O7Na C19H10Br4O5S
373 305 466 960 696 792 648 457 372 809 393 582 774 961 854 670
Absorptive Absorptive Contrast Unknown Contrast/reactive Contrast Unknown Unknown Contrast/reactive Unknown Contrast Absorptive Unknown Contrast Unknown Reactive/contrast
⫹ ⫹⫹ ⫹ ⫹⫹ ⫹ ⫹⫹ ⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹
⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫹ ⫹ 0 ⫹⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹ ⫹⫹⫹ ⫹ ⫹⫹⫹ ⫹⫹
Bright blue Dark blue Dark blue Blue Red Light greenish Light reddish Black Red violet Dark green Red violet Blue Green Dark-blue Violet blueish Dark blue violet
⫹⫹ ⫹⫹ ⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹
Toxic Toxic Safe Safe Safe Safe Safe Toxic Toxic Safe Toxic Safe Safe Safe Safe Safe
ILM ⫽ internal limiting membrane. Grade of staining intensity: 0 ⫽ no staining; ⫹ ⫽ faint staining; ⫹⫹ ⫽ moderate staining; ⫹⫹⫹ ⫽ strong staining. See ref. 13. a Selected for further in vivo experiments in rabbits. b Eosin is a vital stain derivate of fluorescein.
FIGURE 1. After a 360-degree limbal incision, corneal removal was performed, followed by total iris removal. Then, 0.05 mL of 2 concentrations of 0.05% and 0.5% of each dye were placed onto the lens capsule (LC) for 1 minute followed by irrigation with
eyes up to 48 hours after death were examined in the laboratory to determine their integrity. To stain the LC, a 360 – degree-limbal incision was performed for corneal removal, followed by segmented iris removal. Next, 0.05 mL of 0.05% and 0.5% concentrations of each dye were applied to the LC for 1 minute, followed by rinsing with saline solution. An anterior curvilinear capsulorrhexis was carried out with forceps, and selectively removed membranes were sent for histologic analysis after buffered formaldehyde fixation and periodic acid–Schiff staining. The entire anterior segment of the eye was then excised to gain access to the vitreous cavity. After open sky vitrectomy (Accurus; Alcon Laboratories), 0.2 mL of dye solution was injected into the posterior vitreous cavity over the macula for 2 minutes and then was removed by mechanical aspiration, trying to leave the stained ILM clearly visible. A bent 27-gauge needle was used to create the initial ILM tear, followed by grasping with a 23-gauge forceps applied to peel the ILM.
blue, trypan blue, evans blue, light green, sudan black, crystal violet, methyl violet, ICG, and bromophenol blue produced only light (⫹) staining of the LC, whereas fast green and trypan blue promoted moderate (⫹⫹) staining. At the higher concentration of 0.5%, toluidine blue, evans blue, light green, sudan black, crystal violet, fast green, methyl violet, ICG, brilliant blue, and bromophenol blue provided moderate (⫹⫹) and trypan blue provided strong (⫹⫹⫹) staining of the LC (Table 1). Experiments with freshly enucleated porcine eyes with dyes at 0.5% demonstrated no (0) retinal ILM staining after eosin Y exposure, whereas congo red, light green, patent blue, and trypan blue generated only faint (⫹) retinal ILM staining. Exposure of the retinal ILM to methylene blue, toluidine blue, indigo carmine, evans blue, bromophenol blue, and fast green resulted in moderate (⫹⫹) staining, and methyl violet, crystal violet, sudan black, brilliant blue, and ICG yielded strong (⫹⫹⫹) retinal ILM staining in porcine eyes (Figure 2). With most eyes, a continuous 1-piece ILM peeling was difficult to achieve, an expected finding in porcine eyes. Most of the vital dyes used in this investigation at 0.5% stained the vitreous very well. Vitreous examination demonstrated moderate (⫹⫹) staining with congo red, crystal violet, fast green, patent blue, eosin Y, methylene blue, toluidine blue, ICG, and methyl violet; demonstrated strong staining (⫹⫹⫹) with trypan blue, light green, and evans blue; and demonstrated faint (⫹) staining with indigo carmine and sudan black (Figure 3).
RESULTS ● SELECTION OF DYES AND PREPARATION:
All dyes but ICG could be diluted successfully in balanced salt solution; ICG was diluted further in glucose 5% solution. Table 1 summarizes the chemical and staining properties of the vital dyes selected for investigation in this study. Four staining agents, evans blue, congo red, sudan black, and trypan blue, belong to the azo dyes, and 7 dyes, light green, crystal violet, fast green, methyl violet, patent blue, brilliant blue, and bromophenol blue, are in the group of the arylmethane agents. The mechanism of staining of the dyes varies and can involve binding, contrast, or reactivity, although some staining agents such as fast green and evans blue bind to living tissues through unknown mechanisms.
● HISTOLOGIC EVALUATION BY LIGHT MICROSCOPY:
Light microscopy examination revealed the typical fibroblastic characteristics of the anterior LC. No anterior capsule or lens epithelium abnormality in the removed area was observed in any eyes except for surface-parallel intracapsular splits and delamination of the capsulorrhexis resulting from fixation artifact (Figure 4). Gross examination of porcine retinal specimens showed no hemorrhage, inflammation, or detachment. In the cases of methyl violet, toluidine blue, and methylene blue staining, additional deeper layers of the retina also were removed during the ILM peeling maneuver, as cellular elements were present in the histopathologic examination (Figure 4). In addition, the vital dyes crystal violet and sudan black at 0.5% were found to be toxic because they induced severe morphologic damage to the retina, such as producing regions of deformation and destruction of inner
● STAINING OF PORCINE INTERNAL LIMITING MEMBRANE, VITREOUS, AND LENS CAPSULE: The grading of
the staining effect of each dye in porcine eyes is shown in Table 1. The effect in surgically stained LC varied for the different dyes and dye concentrations (Figure 1). With the aid of most dyes, continuous capsulorrhexis was completed successfully, whereas adequate contrast and visibility was reported by the surgeon. Indigo carmine, congo red, brilliant blue, and methylene blue did not stain the LC sufficiently at a concentration of 0.05%; however, patent
saline solution. An anterior curvilinear continuous capsulorrhexis was performed. Surgically stained LC varied for the different dyes and dye concentrations from weak (ⴙ) to strong (ⴙⴙⴙ) coloring. (Top row) LC staining with (Left) 0.05% fast green (FG) and (Right) 0.5% FG induced moderate (ⴙⴙ) staining. (Second row) LC staining with (Left) 0.05% light green (LG) caused faint (ⴙ) staining, whereas (Right) 0.5% LG generated moderate (ⴙⴙ) staining. (Third row) Staining with (Left) 0.05% brilliant blue (BriB) did not color the LC sufficiently, but (Right) the 0.5% concentration promoted moderate (ⴙⴙ) staining. (Bottom row) LC staining with (Left) 0.05% bromophenol blue (BroB) caused faint (ⴙ) staining, whereas (Right) 0.5% LG generated moderate (ⴙⴙ) staining.
VOL. 149, NO. 2
NEW VITAL DYES
FIGURE 2. Histologic specimens of retinas, lens capsules, internal limiting membranes (ILMs), and epiretinal membranes (ERMs). (Top left) Gross examination of specimen of peeled retina of porcine eyes after application of 0.5% bromophenol blue dye. Histologic examination revealed no major toxicity changes in the inner and outer nuclear layer cells or in the nerve fiber layer, whereas successful separation of the ILM was achieved. Arrow separates the ILM peeled area (right) from the unpeeled area (left; hematoxylin and eosin, ⴛ40 magnification). (Top middle) Gross examination of specimen of peeled retina of porcine eyes after application of 0.5% crystal violet dye. The dye was considered to be toxic to the retina because it promoted disorganization and diffuse vacuolization within the nerve fiber layer and ganglion cell layer (hematoxylin and eosin, ⴛ40 magnification). (Top right) In the case of 0.5% methyl violet staining for the ILM peeling maneuver, examination of the ILM disclosed cellular fragments of the retina besides the collagen structure (hematoxylin and eosin, ⴛ25 magnification). (Middle center) Freshly removed ERMs stained with 0.5% evans blue showing strong staining (ⴛ10 magnification). (Middle right) Freshly removed ERMs stained with 0.5% light green (LG) showing moderate staining (hematoxylin and eosin, ⴛ10 magnification). (Bottom left) Gross examination of specimen of peeled retina of porcine eyes after application of 0.5% light green (LG) dye. Histologic examination revealed no signs of retinal cellular damage, and successful separation of the ILM was completed. Arrow separates the ILM peeled area (left) from the unpeeled area (right; hematoxylin and eosin, ⴛ40 magnification). (Bottom middle) Light microscopy examination with periodic acid–Schiff (PAS) staining after application of 0.5% fast green onto the lens capsule (LC) revealing the typical fibroblastic characteristics. No anterior capsule or lens epithelium abnormality in the removed area was observed (hematoxylin and eosin, ⴛ25 magnification). (Bottom right) Light microscopy with PAS staining after application of 0.5% LG onto the LC showing the fibroblastic characteristics of the capsule without any other abnormality (hematoxylin and eosin, ⴛ25 magnification).
FIGURE 3. For evaluation of internal limiting membrane (ILM) staining, the porcine eye cups had the anterior segment removed with a 3-mm incision behind the limbus, and eyes with funduscopic red reflex were included in the experiments. In this figure, ILM staining is shown, which was obtained after placement of 0.05 mL 0.5% of several dyes onto the retinal surface. (Top row) Moderate staining with (Left) fast green FG, and weak staining after (Right) patent blue (PB). (Second row) Strong staining with (Left) indocyanine green (ICG), and faint staining with (Right) congo red (CR). (Third row) Moderate staining with (Left) evans blue (EB) and (Right) indigo carmine (IC). (Fourth row) Moderate staining with (Left) toluidine blue (ToB), whereas (right) trypan blue (TB) caused faint coloring.
nuclear layers and outer nuclear layers, as well as focal areas of damage to photoreceptors. In contrast, eosin Y, evans blue, congo red, brilliant blue, bromophenol blue, indigo carmine, light green, patent blue, ICG, trypan blue, and fast green yielded no major morphologic alterations in porcine retinas, whereas the ILM could be peeled successfully. Histologic examination revealed that these dyes promoted no major toxicity changes in the inner and outer nuclear layer cells, nerve fiber layer, retinal pigment epithelium, and choriocapillaris. Histologic evaluation also demonstrated the successful separation of the ILM in cases of peeling with the aid of evans blue, light green, brilliant blue, bromophenol blue, patent blue, ICG, trypan blue, and fast green, without remnants of retinal cells (Figure 4).
ERM. ERM examination revealed strong (⫹⫹⫹) staining with evans blue and trypan blue; moderate (⫹⫹) staining with 0.5% light green, fast green, brilliant blue, bromophenol blue, and patent blue; and faint (⫹) staining after exposure to indigo carmine and ICG (Table 1). At the lower dose of 0.05%, evans blue, fast green, patent blue, ICG, brilliant blue, and bromophenol blue provided only faint (⫹) staining, and light green and trypan blue produced moderate ERM staining (Figure 4). Histopathologic examination disclosed the highly cellular composition of the ERM with retinal pigment epithelium elements with some connective tissue components. ● STAINING OF LENS CAPSULE, VITREOUS, AND INTERNAL LIMITING MEMBRANE IN HUMAN DONOR EYES:
Most of the dyes stained the LC homogenously, where its edge could be observed, and the circular capsulorrhexis could be performed in all cases. The concentration of 0.05% was enough to produce at least faint (⫹) staining
● STAINING OF FRESHLY EXCISED HUMAN EPIRETINAL MEMBRANE: The results showed that most dyes possess at
least some binding affinity to freshly enucleated human VOL. 149, NO. 2
NEW VITAL DYES
FIGURE 4. Human donor eyes obtained from the eye bank underwent application of vital dyes for staining after corneal and iris removal. An anterior curvilinear capsulorrhexis was carried out with forceps. (Top row) Staining with 0.5% light green (LG) promoted strong lens capsule (LC) coloring. (Middle row) Application of 0.5% bromophenol blue (BroB) induced strong LC staining. (Bottom row) Staining with 0.5% indigo carmine (IC) yielded mild LC coloring.
with clear visualization with most dyes. Exceptionally, indigo carmine produced no staining, and no merit could be found in using this dye at this concentration. However, trypan blue enabled strong staining of the anterior LC. The staining was enhanced as the concentration of the dyes was increased to 0.5%, where mild (⫹) staining occurred with indigo carmine; moderate (⫹⫹) staining occurred with evans blue, fast green, patent blue, ICG, and brilliant blue; and strong (⫹⫹⫹) staining occurred with bromophenol blue, light green, and trypan blue (Figure 5). To achieve ILM staining, the dye solution was injected directly over the macula, where it tended to settle with minimal dispersion (Figure 6). After its removal from the vitreous cavity, the underlying ILM that had been in contact with the dye was stained brightly (⫹⫹⫹) with 0.5% ICG and brilliant blue. However, 0.5% evans blue, light green, and bromophenol blue yielded moderate (⫹⫹) staining of the human ILM, whereas 0.5% indigo carmine, 272
fast green, patent blue, and trypan blue had only a mild effect. The flap of ILM was seen easily, and it could be grasped using intraocular forceps, aided by the distinct contrast between the stained ILM and the unstained retina. The posterior hyaloid of 3 eyes remained attached after limited open sky vitrectomy. The grading of the staining effect of each dye and dye concentration for every intraocular membrane, vitreous, anterior LC, and ILM is shown in Table 2.
DISCUSSION IN RECENT YEARS, DYE-ENHANCED CATARACT AND VIT-
reoretinal surgery with trypan blue and ICG has become the standard method for cases of poor visualization of the LC, ERMs, or ILM. The dyes currently available for intraoperative visualization of tissues in ophthalmology OF
FIGURE 5. Vitreous was removed from fresh enucleated pig eyes and placed in a 5-mL vial for 1 minute with 2 mL of each of the 13 dyes at a concentration of 0.5%. The vitreous then was removed, rinsed thoroughly, and placed on a white background for staining evaluation. Most vital dyes at 0.5% stained the vitreous greatly: moderate coloring with congo red (CR), crystal violet (CV), fast green (FG), patent blue (PB), eosin Y (EY), methylene blue (MB), toluidine blue (ToB), indocyanine green (ICG), and methyl violet (MV); strong staining with trypan blue (TB), light green (LG), and evans blue (EB); and faint staining with indigo carmine (IC), and sudan black (SB).
some of the available dyes may be toxic to ocular tissues.1,3 Therefore, we selected 13 alternative vital dyes to conduct an investigation of their staining potential with regard to the LC, vitreous, ERMs, and ILM.12,14 In this study, we performed a systematic series of experiments to determine the better new vital dyes for eye surgery. The dye concentrations of 0.5% and 0.05% were based on their previous application in other fields of
are: ICG (ICG-Pulsion; Munich, Germany; Indocianina Verde; Ophthalmos, São Paulo, Brazil; and IC-Green; Akorn Inc, Buffalo Grove, Illinois, USA); trypan blue (VisionBlue and MembraneBlue; DORC International, Zuidland, Netherlands), patent blue (Blueron; Fluoron/Geuder, Ulm, Germany), IfCG (Infracyanine green; Serb, Paris, France), and brilliant blue (Brilliant peel; Fluoron/Geuder, Ulm, Germany).1 However, recent research has shown that VOL. 149, NO. 2
NEW VITAL DYES
FIGURE 6. Human donor eyes underwent vitreous and removal for examination of the internal limiting membrane (ILM) staining. (Top left) Bromophenol blue (BroB) at 0.5% promoted moderate staining of the ILM. (Top right) Fast green (FG) at 0.5% generated mild staining of the ILM. (Bottom left) Evans blue (EB) at 0.5% caused moderate staining of the ILM. (Bottom right) Light green (LG) at 0.5% induced moderate staining of the ILM.
medicine and enabled comparison between dyes.12,15 Initially, all new dyes at 0.5% tested in enucleated porcine eyes enabled exclusion of some vital dyes for 3 reasons: weak or no staining of intraocular tissues, any major signs of toxicity in porcine retinal tissue, or the presence of retinal cell fragments in the ILM. When methyl violet, toluidine blue, and methylene blue were used for the ILM peeling maneuver, cellular elements were observed, whereas the dyes crystal violet and sudan black at 0.5% induced retinal damage. Eosin Y and congo red did not produce significant staining to make further testing worth pursuing. These results left the 6 dyes evans blue, brilliant blue, bromophenol blue, indigo carmine, light green, and fast green for further investigation in human donor eyes. These 6 dyes did not cause damage, whereas a considerable intense staining was achieved. During cataract surgery, trypan blue and ICG dyes at concentrations from 0.1% to 0.5% may stain the LC.1 However, trypan blue should be avoided in fertile or pregnant women and in children because of possible teratogenic and mutagenic effects.16 Besides, the dyes have 274
been recently described as possible toxic agents for endothelial cells in certain circumstances.17 Moreover, LC staining in vitrectomized patients may lead to inadvertent staining of the posterior LC because of dye diffusion into the vitreous, thereby obscuring the red reflex. Based on these drawbacks, our group decided to pursue the testing of novel dyes also for LC. We found that most dyes provided some staining of the LC, where the best staining agents for that task may be bromophenol blue, light green, fast green, evans blue, and brilliant blue. In regard to fast green and evans blue, these dyes promoted significant staining of the LC; interestingly, there has been no previous report of their use for LC visualization. In regard to brilliant blue, our study found good staining of the LC at 0.5%, but only a mild effect was achieved with a lower concentration. Other studies have described opposite results concerning brilliant blue; Hisatomi and associates analyzed the effectiveness of brilliant blue for LC visualization at concentrations from 0.001% to 1% in porcine eyes.18 The minimal concentration needed to produce high-quality staining with clear visualization was found to be 0.025%, OF
TABLE 2. Staining Properties in Human Donor Eyes of Novel Selected Dyes Ex Vivo Staining Dye
Concentration Indigo carmine Evans blue Light green SF Fast green Patent bluea Indocyanine greena Trypan bluea Brilliant blue Bromophenol blue
Ex Vivo Staining in Donor Eyes
0.05% 0 ⫹ ⫹⫹ ⫹ ⫹ ⫹ ⫹⫹ ⫹ ⫹
0.5% ⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹
0.05% 0 ⫹ ⫹ ⫹ 0 ⫹⫹ ⫹ ⫹⫹ ⫹⫹
0.5% ⫹ ⫹⫹ ⫹⫹ ⫹ ⫹ ⫹⫹⫹ ⫹ ⫹⫹⫹ ⫹⫹
0.05% ⫹ ⫹⫹ ⫹⫹ ⫹ ⫹ ⫹ ⫹⫹ ⫹⫹ ⫹
0.5% ⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫹ ⫹ ⫹⫹ ⫹⫹ ⫹⫹
0.05% 0 ⫹ ⫹ ⫹ ⫹ ⫹ ⫹⫹ ⫹ ⫹
0.5% ⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹⫹ ⫹⫹ ⫹⫹⫹
ERM ⫽ epiretinal membrane; ILM ⫽ internal limiting membrane. Grade of staining intensity: 0 ⫽ no staining; ⫹ ⫽ faint staining; ⫹⫹ ⫽ moderate staining; ⫹⫹⫹ ⫽ strong staining. See ref. 13. a Not considered novel, but they were included for comparison of staining affinity.
half of the lower dose we chose to study. With regard to bromophenol blue, Schuettauf and associates and Haritoglou and associates reported in surgically extracted LC that among several novel dyes, bromophenol blue demonstrated satisfactory staining characteristics at concentrations of 0.02% to 0.5%.10,19 These results correlate partially with our findings in this study, because bromophenol blue was able to stain the LC at the 2 concentrations investigated, whereas at the lower concentration, only mild staining was found. Haritoglou and associates investigated the staining characteristics of 6 new dyes, including light green, at concentrations from 0.05% to 1.0% for staining of the LC. They reported that light green did not stain the LC sufficiently at concentrations of 0.5% or less.19 In contrast to those results, we found excellent staining with light green at the concentration of 0.5%. Some explanations for different results for this dye include the time of experiments after enucleation, solvents for dye preparation, the presence of salts or chemical substances in the dye, or surgical technique for dye application. Therefore, more clinical studies are necessary before final conclusions can be drawn with regard to those new vital dyes. ICG has become a great tool for ILM identification, and more recently, brilliant blue has been approved in Europe at a concentration of 0.025% for visualization of the ILM.14 Indeed, we observed a strong staining of the ILM with brilliant blue, confirming previous reports.1,3,20 However, a number of studies that suggest remarkable retinal toxicity of ICG and brilliant blue; such observations prompted our group to investigate the staining potential of other dyes. The vital dyes selected as better for the ILM were bromophenol blue, light green, and fast green (Table 1). Fast green produced moderate staining of the porcine ILM and faint staining of the human ILM, whereas light green provided faint staining at the low dose and moderate staining with higher concentration. Nevertheless, mild VOL. 149, NO. 2
NEW VITAL DYES
staining by light green and fast green was enough to allow ILM peeling, and no major damage to the retina was noted. Interestingly, fast green was applied first in vitreoretinal surgery by Sorsby in 1939 as an endovenous agent for identification of breaks in the treatment of retinal detachments.21 In 2005, Jackson and associates examined the staining potential of different types of dyes including fast green in bovine retina.11 Their results showed that fast green produced homogenous but weak staining in some retinal specimens at concentrations as low as 0.05%. With regard to bromophenol blue, in vivo animal studies showed that bromophenol blue at concentrations of 0.5% and 0.02% did not result in any significant retinal toxicity,19 whereas providing enhanced ILM staining, which confirms the good staining of the ILM achieved in our study with bromophenol blue. Therefore, light green, brilliant blue, bromophenol blue, and fast green should be considered in future trials for macular hole surgery. Staining the ERMs may be particularly advantageous because they rarely contain pigment and usually are difficult to recognize during surgery. In this investigation, staining of the ERM obtained from retinal detachment or diabetic retinopathy surgery could be achieved successfully with most dyes, whereas the strongest staining agents were light green, followed by fast green, evans blue, brilliant blue, and bromophenol blue. Human ERM exposed to fast green, evans blue, brilliant blue, and bromophenol blue demonstrated moderate staining with the higher dose of 0.5% and faint staining at 0.05%. Haritoglou and associates investigated staining characteristics of ERMs with dye concentrations of 1.0%, 0.5%, 0.2%, and 0.05%, and found that among 6 stains, including light green and E68, bromophenol blue stained the ERM best.19 There is not much information to date concerning the staining capacity of other dyes such as fast green and evans blue. However, it can be assumed that evans blue may possess chemical IN
related to ILM removal has been performed already.22,23 In addition, porcine eyes are more readily available, and the short interval between animal death and experiments may reduce the incidence of autolysis in the retina.24 Nevertheless, the results of this study may not be totally applicable to humans, because pigs may have different subtypes of ILM collagen. In another branch of our research, we recently determined the biocompatibility of the 6 dyes investigated thoroughly in this study. Evans blue and light green induced some signs of functional and morphologic retinal damage in rabbits.25 Some strengths of the present study include a systematic investigation of several dyes, concern for obtaining preliminary data in animals before human testing, and better quality of data of animal studies in comparison with cell culture models. In summary, this study showed that brilliant blue, bromophenol blue, fast green, and light green stain the ILM well, and they therefore are good alternative dyes to ICG, and that ERMs can be stained successfully by light green, fast green, evans blue, brilliant blue, and bromophenol blue. All dyes stained the vitreous and may be considered for human use in the future. Future studies of these vital dyes may improve the indications for retinal and vitreous staining, may increase the precision of surgical techniques, and may enhance the discrimination between the several intravitreous targets.26
moieties that bind to ERMs, because they also are azo dyes like trypan blue, which is an excellent dye for the ERM. Vitreous staining was achieved with all dyes tested in this study in porcine eyes. Trypan blue, light green, and evans blue led to strong vitreous staining (⫹⫹⫹), whereas toluidine blue, congo red, crystal violet, bromophenol blue, brilliant blue, fast green, methyl violet, ICG, patent blue, eosin Y, and methylene blue caused moderate staining. Our results revealed that all 6 dyes stained to some degree the vitreous tissue in human cadaveric eyes, although indigo carmine to a lesser extent. We postulate that the affinity of the dyes for the vitreous is the result of the hydrophilic characteristic of the dyes in relation to the high water content of the vitreous. Considering future applications, there are 2 methods to choose a dye. First, one may consider applying dyes that also stain both the ERM and ILM, because costs and product registering can be reduced. However, when one dye stains more than one intraocular target, it can lead to intraoperative confusion as to which membrane has been stained. In this regard, it can be proposed to restrict intraoperative use of dyes to those that stain only the vitreous and not other targeted tissues. In our current research, we elected to use porcine eyes for several reasons. First, the morphologic features of the ILM in pigs are similar to those of humans, and research
THE AUTHORS INDICATE NO FINANCIAL SUPPORT OR FINANCIAL CONFLICT OF INTEREST. INVOLVED IN DESIGN AND conduct of the study (E.B.R., F.M.P., E.d.P.F.C., E.D., M.M.Jr., O.M., G.B.M., V.S., M.E.F.); Collection, management, analysis, and interpretation of the data (E.B.R., F.M.P., E.d.P.F.C., M.M., M.M.Jr., C.H.M., A.B.D., V.S., M.E.F.); and preparation, review, or approval of the manuscript (E.B.R., O.M., E.D., G.M.L., E.d.P.F.C., M.M., C.H.M., A.B.D., M.E.F.). The study was approved by the ethics committee of the Federal University of São Paulo, São Paulo, Brazil. Care of the animals was conducted according to Association for Research in Vision and Ophthalmology guidelines. The authors thank the German Academic Exchange Service (DAAD), Fehr Foundation (Germany), Pan-American Ophthalmological Foundation (PAOF), and Fundaçã o de Amparo à Pesquisa do Estado de São Paulo (FAPESP).
8. Enaida H, Hisatomi T, Hata Y, et al. Brilliant blue G selectively stains the internal limiting membrane/brilliant blue G-assisted membrane peeling. Retina 2006;26:631– 636. 9. Mennel S, Thumann G, Peter S, Meyer CH, Kroll P. Influence of vital dyes on the function of the outer bloodretinal barrier in vitro. Klin Monatsbl Augenheilkd 2006; 223:568 –576. 10. Schuettauf F, Haritoglou C, May CA, et al. Administration of novel dyes for intraocular surgery: an in vivo toxicity animal study. Invest Ophthalmol Vis Sci 2006;47:3573– 3578. 11. Jackson TL, Griffin L, Vote B, Hillenkamp J, Marshall J. An experimental method for testing novel retinal vital stains. Exp Eye Res 2005;81:446 – 454. 12. Kiernan JA. Classification and naming of dyes, stains and fluorochromes. Biotech Histochem 2001;76:261–278. 13. Pandey SK, Werner L, Escobar-Gomez M, Roig-Melo EA, Apple DJ. Dye-enhanced cataract surgery. Part 1: anterior capsule staining for capsulorrhexis in advanced/white cataract. J Cataract Refract Surg 2000;26:1052–1059. 14. Rodrigues EB, Maia M, Meyer CH, Penha FM, Dib E, Farah ME. Vital dyes for chromovitrectomy. Curr Opin Ophthalmol 2007;18:179 –187.
REFERENCES 1. Rodrigues EB, Costa EF, Penha FM, et al. The use of vital dyes in ocular surgery. Surv Ophthalmol 2009; 54:576 – 617. 2. Burk SE, Da Mata AP, Snyder ME, Rosa RH Jr, Foster RE. Indocyanine green-assisted peeling of the retinal internal limiting membrane. Ophthalmology 2000;107:2010 –2014. 3. Rodrigues EB, Meyer CH, Kroll P. Chromovitrectomy: a new field in vitreoretinal surgery. Graefes Arch Clin Exp Ophthalmol 2005;243:291–293. 4. Gandorfer A, Haritoglou C, Gandorfer A, Kampik A. Retinal damage from indocyanine green in experimental macular surgery. Invest Ophthalmol Vis Sci 2003;44:316 –323. 5. Maia M, Margalit E, Tso MOM, et al. Effects of intravitreal indocyanine green injection in rabbits. Retina 2004;24: 69 –79. 6. Rodrigues EB, Meyer CH, Farah ME, Kroll P. Mechanisms of intravitreal retinal toxicity to indocyanine green dye: implications for chromovitrectomy. Retina 2007;27:958 –970. 7. Mennel S, Meyer CH, Tietjen A, Rodrigues EB, Schmidt JC. Patent blue: a novel vital dye in vitreoretinal surgery. Ophthalmologica 2006;220:190 –193.
21. Sorsby A. Vital staining of the retina: preliminary clinical note. Br J Ophthalmol 1939;23:20 –24. 22. Janknecht P, Feltgen N, Wesendahl T, et al. Internal limiting membrane ablation in pig eyes with the Er:YAG laser under perfluorodecalin. Graefes Arch Clin Exp Ophthalmol 2001;239:705–711. 23. Czajka MP, McCuen BW, Cummings TJ, et al. Effects of indocyanine green on the retina and retinal pigment epithelium in a porcine model of retinal hole. Retina 2004;24:275– 282. 24. Grisanti S, Szurman P, Gelisken F, Aisenbrey S, OficjalskaMlynzak J, Bartz-Schmidt KU. Histological findings in experimental macular surgery with indocyanine green. Invest Ophthalmol Vis Sci 2004;45:282–286. 25. Rodrigues EB, Penha FM, Costa EF, et al. Pre-clinical retinal biocompatibility of novel dyes for chromovitrectomy. Retina 2009; 29:497–510. 26. Costa EP, Rodrigues EB, Farah ME, et al. Vital dyes and light sources for chromovitrectomy: comparative assessment of osmolarity, pH, and spectrophotometry. Invest Ophthalmol Vis Sci 2009;50:385–391.
15. Haigh PI, Lucci A, Turner RR, Bostick PJ, Krasne DL, Stern SL, Morton DL. Carbon dye histologically confirms the identity of sentinel lymph nodes in cutaneous melanoma. Cancer 2001;92:535–541. 16. Melles GRJ, Waard PWT, Pameyer JH, Beekhuis WH. Trypan blue capsule staining in cataract surgery. J Cataract Refract Surg 1999;24:7–9. 17. Chang YS, Tseng SY, Tseng SH, Chen YT, Hsiao JH. Comparison of dyes for cataract surgery. Part 1: cytotoxicity to corneal endothelial cells in a rabbit model. J Cataract Refract Surg 2005;31:792–798. 18. Hisatomi T, Enaida H, Matsumoto H, et al. Staining ability and biocompatibility of brilliant blue G: preclinical study of brilliant blue G as an adjunct for capsular staining. Arch Ophthalmol 2006;124:514 –519. 19. Haritoglou C, Yu A, Freyer W, et al. An evaluation of novel vital dyes for intraocular surgery. Invest Ophthalmol Vis Sci 2005;46:3315–3322. 20. Enaida H, Hisatomi T, Goto Y, et al. Preclinical investigation of internal limiting membrane staining and peeling using intravitreal brilliant blue G. Retina 2006;26:623– 630.
VOL. 149, NO. 2
NEW VITAL DYES