Bupivacaine for topical eye anesthesia


Joseph Eldor, MD

Department of Anesthesia

Kaplan Medical Center,

Rehovot, Israel





Albert Niemann (1880-1921) separated the alkaloid cocaine from the dried leaves of the coca plant in 1860. He studied the white powder and named it cocaine, also noting the temporary numbing effect the compound had on his tongue. During the 1880s in Vienna, Austria, Sigmund Freud (1856-1939) studied cocaine as a treatment for morphine addiction. Freud suggested the possible use of cocaine as a local anesthetic to Viennese colleagues Leopold Königstein, a professor of ophthalmology (the medical study of the eye and diseases of the eye), and Carl Koller (1857-1944), a young ophthalmologist (doctor specializing in eye diseases).

Koller experimented on animals and then presented his findings to the Congress of Ophthalmology in Heidelberg, Germany, in 1884. He demonstrated the successful use of cocaine as a local anesthetic during eye surgery. Koller's findings were accepted enthusiastically. Koller himself emigrated to the United States in 1888 and established practice in New York City, where he died in 1944. Cocaine was used widely for ophthalmological procedures until it was discovered that it causes damage to the cornea.


Oxybuprocaine corneal toxicity


Topical ocular anesthesia has been part of ophthalmology for more than a century. The most commonly used drugs today are proparacaine, tetracaine, benoxinate (oxybuprocaine) cocaine and lidocaine. Although generally well tolerated, all these can be toxic, particularly when abused. The most common toxicities are to the ocular surface, but abuse can cause deep corneal infiltrates, ulceration and even perforation. Fortunately, systemic side effects are rare. Cocaine is unique for its higher incidence of systemic side effects and high abuse potential, both of which impede its clinical use. When used appropriately, all these drugs are remarkably safe. They are generally not prescribed for home use, as prolonged abuse of these drugs can be expected to result in serious complications (1).

Nam et al. (2) compared changes in human corneal thickness after the instillation of proparacaine with those after oxybuprocaine instillation with time over a period of 10 minutes. Eighteen healthy young participants were recruited. Proparacaine was used in the right eye and oxybuprocaine in the left. Right and left baseline corneal thicknesses were measured every 30 seconds for 10 minutes using a noncontact specular microscope by 1 observer. Baseline corneal thickness was defined as the average of all values taken over 10 minutes. Changes in corneal thickness were measured every 20 seconds for 10 minutes after the administration of 1 drop of 0.5% proparacaine onto the right cornea and 1 drop of 0.4% oxybuprocaine onto the left cornea. Mean baseline right cornea thickness was 531 +/- 45 microm, and that of the left cornea was 531 +/- 42 microm. The corneal thickness after proparacaine increased by 8.6 microm ( approximately 4.5-12.6 microm, 95% CI) and then returned to baseline within 80 seconds. Corneal thickness after applying oxybuprocaine increased by 7.7 microm (3.6-11.2 microm, 95% CI) and then returned to baseline within 80 seconds. There was a second transient increase about 5 minutes later after proparacaine instillation but no additional transient increase after oxybuprocaine instillation. Oxybuprocaine is similar to proparacaine in terms of the severity of its effect on corneal thickness. Corneal thickness instability may occur for 5 minutes after proparacaine administration. Changes in corneal thickness after topical anesthetic instillation should be considered when performing measurements for refractive surgery or central corneal thickness in glaucoma patients.

Asensio et al. (3) determined changes in corneal thickness after topical anesthesia. Corneal thickness was measured before and 3 minutes after administration of two drops of oxybuprocaine 0.4% to 26 patients (26 eyes). They analyzed the corneal thickness of a control group, which was made up of 26 patients (26 eyes) before and 3 minutes after administration of two drops of saline solution. Corneal thickness was measured with the Orbscan Topography System II (Bausch Lomb Surg., Barcelona). Variations higher than +/- 10 microm were found following the instillation of 2 oxybuprocaine eye drops in eight eyes (30.76%) at the inferonasal cornea, in six eyes (23.08%) at the superotemporal, temporal and inferotemporal cornea, in five eyes (19.23%) at the nasal cornea, in three eyes (11.53%) at the central cornea, and in two eyes (7.69%) at the superonasal cornea. Nevertheless, no significant differences in the mean corneal thickness at each corneal location between the first and the second corneal thickness measurements were found in anesthetized eyes. Some individuals can present important increases and decreases in corneal thickness values after anesthetic eye drops. This effect of anesthetic eye drops must be considered by refractive surgeons when carrying out preoperative laser in situ keratomileusis corneal thickness measurements.

The effect of topical anaesthetic ointments (4% cocaine, 4% xylocaine, 0.5% proparacaine, 0.2% oxybuprocaine) on the corneal epithelium of rabbits was examined using a scanning electron microscope (4). Even after a single application, the more toxic effect of cocaine compared to the other topical anaesthetics was evident. Cocaine caused disruption of both the plasma membrane and the cytoplasm. After a single application, the other preparations caused a marked decrease in the microvilli and microplicae, disruption of the intercellular spaces and the prominence of the cell nucleus which under normal condition is not visible. Repeated applications caused regular cell desquamation and damage to the plasma membrane and cytoplasm. The damage also affected several cell layers. The cell reactions described are clearly a consequence of the topical anaesthetics, as the ointment base itself produced no essential cell damage. The scanning electron microscope findings were supported by results from the transmission electron microscope.

The effect of different local anesthetics (Cocain 4%, Lidocaine 2%, Proparacain) on the corneal epithelium in rabbits was examined under scanning electron microscope (5). The experiment was divided into three groups. Group 1 received one application of two drops of the given local anesthetic for a reaction time of 5 minutes. Group 2 received two drops of the given anesthetic after 0, 5 and 10 minutes. The cornea was excised after 15 minutes. Group 3 were measured after a single application of Proparacain using a Schiötz or hand applanation tonometer according to Draeger. After a single dose of a local anesthetic principally the same changes in the surface of the cornea were observed with all the preparations used: a distinct decrease in the number of microvilli and microplicae, disruption of the intercellular spaces and the prominence of the cell nuclei which under normal conditions are not visible. After several applications the greater toxicity of Cocain compared with the other preparations was clearly seen through the disruption of the plasma membrane and the cytoplasm. The damage effected several layers of cells. Tonometry when correctly performed causes no additional damage to the cell surface.- The effect of local anesthetics on the cell membrane can only take place after the disruption of the tear film. The results emphasize that local anesthetics should only be applied when absolutely essential.

With double-masking procedures, the dose-response curves for 0.1, 0.2, and 0.4% benoxinate and 0.125, 0.25, and 0.50% proparacaine hydrochloride were determined by monitoring changes in corneal touch threshold after applying each anesthetic (6). The level of corneal anesthesia necessary for applanation tonometry was also determined. The maximum increase in threshold that could be measured following instillation of 50 microliter of the drug was 200 mg/mm2. All 6 anesthetic solutions produced this amount of decreased corneal sensitivity. Recovery from the anesthetic was exponential for all concentrations; however, the lower doses had the shortest duration. For applanation tonometry, the corneal threshold for touch must be 75 mg/mm2 or higher. It was concluded that a quarter to a half of the commonly used anesthetic dose is sufficient for routine tonometric evaluation.

Sun et al. (7) compared the onset time, duration of action, corneal toxicity, and corneal epithelial healing time of 4 topical anesthetic agents in rabbits. Fifty-six rabbits were treated with 4 topical anesthetics (bupivacaine, lidocaine, procaine, and benzocaine) at different concentrations and different pH of solutions. Corneal sensation, corneal toxicity, and corneal epithelial healing time were measured. The onset time of all 4 anesthetic agents was within 1 minute; however, bupivacaine and lidocaine produced significantly longer action than procaine or benzocaine (P < .05). Buffered bupivacaine and lidocaine had a significantly longer anesthetic effect than that of the nonbuffered solutions (P < .05). No significant effect on corneal epithelial healing time or corneal toxicity was observed. Topical bupivacaine and lidocaine had a longer anesthetic effect, particularly in buffered solutions. No significant corneal toxicity was observed.

Judge et al. (8) determined the relative corneal endothelial toxicities of the following topical anesthetic agents: bupivacaine HCl 0.75%, unpreserved lidocaine HCl 4%, proparacaine HCl 0.5%, and tetracaine HCl 0.5%. The experiment was conducted using pigmented rabbits. Approximately nine animals each were randomly assigned to eight groups. Right eyes received injections of 0.2 ml of one of the four anesthetic agents at one of two concentrations and left eyes received injections of 0.2 ml of balanced salt solution. Corneal thickness and clarity were measured before surgery and on postoperative days 1, 3, and 7. A statistically significant increase (P < 0.05) in corneal thickness and opacification over preoperative measurements was noted with injections of bupivacaine, lidocaine, and proparacaine, controlling for changes occurring in control eyes from surgery alone. Proparacaine was statistically more toxic than were the others. The toxicity of tetracaine was statistically indistinguishable from balanced salt solution, although mild toxicity was evident clinically. Injection of 1:10 dilutions of the same anesthetic agents failed to produce a statistically significant increase in corneal thickness or opacification on any postoperative examination. Anterior chamber injection of bupivacaine HCl 0.75%, unpreserved lidocaine HCl 4%, and proparacaine HCl 0.5% produces corneal thickening and opacification that is clinically and statistically significant. Tetracaine HCl 0.5% injection produces corneal thickening and opacification that is clinically apparent in some eyes but statistically insignificant. Ophthalmic surgeons should be aware of the potential for endothelial cell injury if anesthetic agents enter or are injected into the eye during cataract surgery in the concentrations supplied commercially.

Gao et al. (9) studied the effects of Complex Tropicamide (0.5% Tropicamide and 0.5% phenylephrine HCI) and Saline solution on corneal thickness in adult myopic patients with the Orbscan II system in a prospective, nonrandomized, clinical trial. The thinnest pachymetry of the cornea was obtained before and 1.5 hours after administration of three drops of Complex Tropicamide to the left eyes of 58 patients (58 eyes) and Saline solution to their 31 right eyes, respectively. The corneal thickness of the other 27 right eyes before and 1.5 hours after eyelid closure without exposure to eye drops was used as the control group. The thinnest pachymetry of the cornea was significantly higher after exposure to eye drops in the Tropicamide group (23.36 +/- 15.01 microm; t = -11.855, P < 0.001). Similar findings were also noted in the Saline group (7.13 +/- 8.11 microm, t = -4.894, P < 0.001). The difference between the two groups was also significant (t = 6.737, P < 0.001). There were no statistically changes in corneal thickness in control group. The drops tested have no effect on the location of the thinnest corneal site and its distance form the visual axis. Eye drops including Saline solutions may have significant effects on the corneal thickness in myopia, and this may have implications for corneal refractive surgery.


Bupivacaine topical eye anesthesia


Guzey et al. (10) investigated the direct toxic effects of bupivacaine HCl 0.5% and lidocaine HCl 2%, two commonly used injectable local anesthetic agents, on the corneal endothelium when applied intracamerally. Two groups were formed, each consisting of 12 pigmented rabbits, and 0.2 ml of the anesthetic agent were injected intracamerally into the right eyes. The central corneal thicknesses and corneal clarities were evaluated preoperatively and at 3, 6, 9, 12 h and 1, 3, 7 days postoperatively. While the central corneal thicknesses were evaluated by ultrasonic pachymetry, the corneal opacification scored between 0 and 3 was assessed by biomicroscopic examination and photographs. Both bupivacaine and lidocaine caused corneal thickening in the 3- to 12-hour measurements. In addition, there was significant corneal opacification in both groups in the 3-hour and 3-day measurements. The corneal thickening and corneal opacification determined during 3- and 6-hour measurements in the eyes which received intracameral bupivacaine were significantly higher than those determined in the lidocaine-injected group. In both groups, the corneal thickness and corneal clarity scores returned to the preoperative values on the 1st and 7th days, respectively. When applied into the anterior chamber at the concentrations supplied commercially, both bupivacaine and lidocaine cause statistically significant corneal thickening and clinically significant corneal opacification. It should be noted that the injection of these agents into the anterior chamber during the operation at the concentrations supplied commercially may be a potential risk factor for endothelial injury.

Bupivacaine is a local ocular anesthetic with a long duration of action when administered by retrobulbar injection. To determine the potential for the use of bupivacaine as a topical ocular anesthetic, the onset and duration of action and toxicity of various concentrations of bupivacaine were studied after instillation in rabbit eyes. The onset and duration of action were not significantly different from that of topical 0.5% proparacaine. Increasing the pH of the bupivacaine solution from 5.7 to 6.5 nearly doubled the duration of action, but the increase was not sufficient to be clinically important. Slit lamp biomicroscopic examination and scanning electron microscopy showed that bupivacaine was less toxic to the corneal epithelium than 0.5% proparacaine. Healing after keratectomy was significantly more rapid in eyes treated with 0.75% bupivacaine, compared with eyes treated with 0.5% proparacaine. These results suggest that bupivacaine may be less toxic to the corneal epithelium than proparacaine and could be clinically useful for topical ocular anesthesia, particularly if pharmacologic modifications can increase the duration of anesthesia provided by this drug (11).

Anderson et al. (12) examined whether intracameral bupivacaine hydrochloride 0.5% is as effective as lidocaine hydrochloride 1.0% in controlling discomfort of patients during phacoemulsification and posterior chamber intraocular lens implantation. In rabbits, corneal endothelial cell function, ultrastructure, and viability were evaluated after in vitro perfusion of bupivacaine 0.5%. In a double-masked, controlled trial, 48 eyes of 48 patients with uncomplicated age-related cataract were randomly assigned to receive bupivacaine 0.5% or lidocaine 1.0% intracamerally before phacoemulsification with a posterior chamber intraocular lens. Outcome measures such as pain, visual acuity, amount of sedation, length of surgery, pupil size, intraocular pressure, corneal clarity, and anterior chamber reaction were compared. In laboratory studies, paired rabbit corneas were evaluated by endothelial cell perfusion with either bupivacaine 0.5%, bupivacaine 0.5% and glutathione bicarbonate Ringer solution in a 1:1 ratio or bupivacaine 0.5% buffered to a pH of 7.0. The paired control corneas were perfused with glutathione bicarbonate Ringer solution and rates of corneal swelling were determined. Cell ultrastructure and viability were also evaluated. In the randomized trial, there was no significant difference in the pain patients had during surgery or in the early or late postoperative period. No statistically significant difference was seen between the two groups in terms of pupil size, intraocular pressure, corneal edema, anterior chamber reaction, or visual acuity immediately after the operation or on postoperative day 1. Paired rabbit corneas perfused with bupivacaine 0.5% and bupivacaine 0.5% buffered to a pH of 7.0 swelled significantly (P<.001, P = .009, respectively), and had corneal endothelial cell damage. Dilution of the bupivacaine 1:1 with glutathione bicarbonate Ringer solution prevented corneal edema and damage to the corneal endothelium. Endothelial cell viability was also decreased after perfusion of bupivacaine 0.5% (P<.001). Clinically, bupivacaine 0.5% is as effective as lidocaine 1.0% for anesthesia during phacoemulsification and posterior chamber intraocular lens implantation. However, in vitro perfusion of bupivacaine 0.5% damaged the corneal endothelium of rabbits except when the drug was diluted 1:1 with glutathione bicarbonate Ringer solution. Surgeons who use 0.2 to 0.5 ml of intracameral bupivacaine 0.5% should be aware of its potential to cause endothelial cell damage because of its lipid solubility. The bupivacaine 0.5% should be diluted at least 1:1 with balanced salt solution before intracameral injection, followed immediately by phacoemulsification. The surgeon should ensure that the bupivacaine 0.5% is nonpreserved and packaged in single-use vials or flip-top containers.


Antibacterial activity of Bupivacaine topical eye anesthesia


The antibacterial effect of topical anesthetics may lead to false-negative cultures from corneal specimens of bacterial keratitis. Pelosini et al. (13) compared the antibacterial effect of 3 unpreserved topical anesthetics to indicate the most appropriate agent for corneal scrapes. Four bacterial strains (Staphylococcus epidermidis, Staphylococcus aureus, Pseudomonas aeruginosa, and Streptococcus pneumoniae) derived from the most frequently isolated microorganisms from corneal ulcers were cultured from stored control stocks and clinical specimens. These strains were used to determine the minimum inhibitory concentration (MIC) of 3 preservative-free anesthetic eye drops: proxymetacaine 0.5%, oxybuprocaine 0.4%, and tetracaine 1%. There was no inhibition of growth seen with proxymetacaine 0.5% (5000 microg/mL) with any of the organisms except S. epidermidis, which demonstrated a MIC of 2500 microg/mL (equivalent to a dilution of (1/2)). Tetracaine 1% (10,000 microg/mL) produced a MIC ranging between 625 and 1250 microg/mL, inhibiting all 4 strains at the commercially available dilution. Oxybuprocaine 0.4% (4000 microg/mL) resulted to be the second most inhibitory preparation with a MIC ranging between 1000 and 2000 microg/mL. Currently used preservative-free topical anesthetics differ in bacterial growth inhibition. This in vitro study showed that proxymetacaine 0.5% is the least inhibitory on bacterial growth and therefore the most appropriate to be used before corneal scrapes.

Dantas et al. (14) investigated the antibacterial activity of topical anesthetic solutions and their preservatives individually in vitro, to determine involvement with bacterial growth inhibition. Proparacaine and tetracaine (in concentrations of 0.125%, 0.25%, and 0.50%), edetate disodium (EDTA), benzalkonium chloride, EDTA + benzalkonium chloride, and sterile saline solution were used. Five microliters of each solution were applied to standard filter paper disks and placed in Mueller-Hinton agar previously inoculated with known strains of Pseudomonas aeruginosa and Staphylococcus aureus. Zones of growth inhibition were measured 24 hours later and analyzed. There were no zones of inhibition in the agar inoculated with P. aeruginosa to all tested solutions. Benzalkonium chloride alone and associated with EDTA inhibited growth of S. aureus. All other solutions did not inhibit S. aureus. Preservative-free anesthetic solutions seemed not to interfere with bacterial development in culture media. Benzalkonium chloride alone and associated with EDTA inhibited development of gram positive bacteria, S. aureus, but did not inhibit P. aeruginosa.

Topical anesthetics are commonly used prior to obtaining bacterial cultures in ulcerative keratitis. Mullin and Rubinfeld (15) performed an in vitro study designed to test both the bacteriostatic and bactericidal effects of commercially available preserved topical anesthetic agents. Proparacaine, tetracaine, cocaine, and sterile water solutions were applied to filter paper disks, which were then placed on Mueller-Hinton agar plates that had previously been inoculated with known quantities of Pseudomonas aeruginosa and Staphylococcus aureus. After 24 h of incubation, zones of inhibition were measured and recorded. Proparacaine strongly inhibited the growth of S. aureus at all concentrations (0.5%, 0.25%, 0.125%) and inhibited growth of P. aeruginosa at 0.5% and 0.25% but not at 0.125% concentration. Tetracaine also inhibited S. aureus at 0.5% and inhibited P. aeruginosa at 0.5% and 0.25% concentrations. Cocaine exhibited no inhibition of S. aureus and exhibited mild inhibition of P. aeruginosa growth only at the 4% concentration. The in vitro antibacterial effect of topical anesthetics suggests one possible reason why bacterial culture yields in clinical ulcerative keratitis are suboptimal. It is proposed that clinicians consider the use of a 1% or 2% cocaine solution instead of standard commercial topical anesthetics in the management of individual cases of ulcerative keratitis and in future clinical bacterial keratitis studies.

The antibacterial activities of local anesthetics are recognized. Gocmen et al. (16) investigated in vitro the activity of topical local anesthetic ointments at clinical doses. The activity of two different local anesthetic ointments including lidocaine 5% and lidocaine/prilocaine 2.5% was tested against Staphylococcus aureus, Staphylococcus epidermidis, Escherichia coli, Pseudomonas aeruginosa, Streptococcus pyogenes and Enterococcus faecalis by the disc-diffusion method. Sterile discs containing topical local anesthetic drugs were prepared taking into account the doses of ointments used in clinical practice. The validity of the methodology was confirmed using topical antibacterial mupirocin. The inhibition zones of the discs were measured. Mupirocin inhibited all the bacteria. Both local anesthetic ointments were found to be most effective on E. coli, whereas they had no effects on P. aeruginosa. Lidocaine 5% revealed antibacterial activity against S. aureus, S. epidermidis, E. coli, S. pyogenes and E. faecalis, but lidocaine/prilocaine 2.5% showed no activity on E. faecalis and inhibited S. pyogenes only at double doses. It was also observed that the antibacterial activity was in a dose-dependent manner. In the light of these findings, it might be concluded that topical local anesthetic ointments in routine settings may have a preventive role against some bacteria.




1. McGee HT, Fraunfelder FW. Toxicities of topical ophthalmic anesthetics. Expert Opin Drug Saf. 2007 Nov;6(6):637-40.

2. Nam SM, Lee HK, Kim EK, Seo KY. Comparison of corneal thickness after the instillation of topical anesthetics: proparacaine versus oxybuprocaine. Cornea. 2006 Jan;25(1):51-4.

3. Asensio I, Rahhal SM, Alonso L, Palanca-Sanfrancisco JM, Sanchis-Gimeno JA. Corneal thickness values before and after oxybuprocaine 0.4% eye drops. Cornea. 2003 Aug;22(6):527-32.

4. Brewitt H, Bonatz E, Honegger H. Morphological changes of the corneal epithelium after application of topical anaesthetic ointments. Ophthalmologica. 1980;180(4):198-206.

5. Brewitt H, Honegger H. The influence of local anesthetics on corneal epithelium. A scanning electron microscopic study. Klin Monatsbl Augenheilkd. 1978 Sep;173(3):347-54.

6. Polse KA, Keener RJ, Jauregui MJ. Dose-response effects of corneal anesthetics. Am J Optom Physiol Opt. 1978 Jan;55(1):8-14.

7. Sun R, Hamilton RC, Gimbel HV. Comparison of 4 topical anesthetic agents for effect and corneal toxicity in rabbits. J Cataract Refract Surg. 1999 Sep;25(9):1232-6

8. Judge AJ, Najafi K, Lee DA, Miller KM. Corneal endothelial toxicity of topical anesthesia. Ophthalmology. 1997 Sep;104(9):1373-9.

9. Gao L, Fan H, Cheng AC, Wang Z, Lam DS. The effects of eye drops on corneal thickness in adult myopia. Cornea. 2006 May;25(4):404-7.

10. Guzey M, Satici A, Dogan Z, Karadede S. The effects of bupivacaine and lidocaine on the corneal endothelium when applied into the anterior chamber at the concentrations supplied commercially. Ophthalmologica. 2002 Mar-Apr;216(2):113-7.

11. Liu JC, Steinemann TL, McDonald MB, Thompson HW, Beuerman RW. Topical bupivacaine and proparacaine: a comparison of toxicity, onset of action, and duration of action. Cornea. 1993 May;12(3):228-32.

12. Anderson NJ, Nath R, Anderson CJ, Edelhauser HF. Comparison of preservative-free bupivacaine vs. lidocaine for intracameral anesthesia: a randomized clinical trial and in vitro analysis. Am J Ophthalmol. 1999 Apr;127(4):393-402.

13. Pelosini L, Treffene S, Hollick EJ. Antibacterial activity of preservative-free topical anesthetic drops in current use in ophthalmology departments. Cornea. 2009 Jan;28(1):58-61.

14. Dantas PE, Uesugui E, Nishiwaki-Dantas MC, Mimica LJ. Antibacterial activity of anesthetic solutions and preservatives: an in vitro comparative study. Cornea. 2000 May;19(3):353-4.

15. Mullin GS, Rubinfeld RS. The antibacterial activity of topical anesthetics. Cornea. 1997 Nov;16(6):662-5.

16. Sedef Gocmen J, Buyukkocak U, Caglayan O, Aksoy A. In vitro antibacterial effects of topical local anesthetics. J Dermatolog Treat. 2008;19(6):351-3