- Email: [email protected]
Experimental study of tetrodotoxin, a long-acting topical anesthetic
Experimental Study of Tetrodotoxin, a Long-acting Topical Anesthetic DANIEL M. SCHWARTZ, MD, HOWARD 1. FIELDS, MD, PHD, KEITH G. DUNCAN, PHD, JACQUE 1. DUNCAN, MD, AND MATTHEW R. JONES, MD To determine the effectiveness and toxicity of tetrodotoxin for use as a long-acting topical anesthetic. l METHODS: Four groups of six rabbits each received a 400~1 aliquot of either tetrodotoxin in one of three concentrations ( 10 mM, 1 mM, or 0.1 mM) or proparacaine 0.5% into the inferior conjunctival cul-de-sac of one eye, with the fellow eye of each rabbit receiving 40 ~1 of a 60-mM, pH 4.3 sodium citrate vehicle as a control. Cornea1 sensation was tested for up to 8 hours after administration of drugs, and response was noted by no blink, partial blink without full eyelid closure, and full blink. Slit#lamp examination at 12 and 24 hours after administration and pachymetry before and 24 hours after administration were performed to detect cornea1 toxicity. l RESULTS: Rabbits receiving all three concentrae tions of tetrodotoxin did not demonstrate any ocular irritation, cornea1 thickening, or signs of systemic toxicity. At a dose of 10 mM, tetrodotoxin produced an anesthetic effect lasting up to 8 hours. At I mM, tetrodotoxin was an effective but shortereacting anesthetic. At 0.1 mM, tetrodotox+ in had no significant anesthetic effect. ProparaCaine-treated rabbits initially were anesthetic, but l
PURPOSE:
Accepted for publication July 24, 1997. From the Department of Ophthalmology (Drs Schwartz, K. Duncan. I. Duncan, and Jones) and the-Departments of Neurology and Physiology (Dr Fields), Universitv of California San Francisco Medical Center. San Francisco, California. Supported in part by an unrestricted grant and core grant EY02162 from Research to Prevent Blindness, Inc, New York, New York. Drs Schwartz and Fields have filed a patent for use of tetrodotoxin. Reprint requests to Daniel M. Schwartz, MD, Department of Ophthalmology, UCSF, 10 Kickham St, Box 0730, San Francisco, CA 941430730; fax: (415) 476-0336; e-mail: [email protected]
0002-9394/98/$19.00
0
1998
BY ELSEVIER SCIENCE
this effect was largely gone by 1 hour and complete0 ly gone by 3 hours. l CONCLUSIONS: Tetrodotoxin is a long-acting topical anesthetic in the rabbit cornea. Although additional toxicity studies are required, tetrodoe toxin may provide an effective, long-lasting topical anesthetic for use in pain control after cornea1 procedures such as photorefractive keratectomy. (Am J Ophthalmol 1998;125:481-487. 0 1998 by Elsevier Science Inc. All rights reserved.)
P
OSTOPERATIVE
PAIN
CAN
BE A DISABLING
COM-
plication of photorefractive keratectomy. Current management with bandage contact lens, nonsteroidal anti-inflammatory agents, and oral analgesics mitigates, but does not eliminate, the discomfort in most patients. 1-3Topical anesthetics have been used to reduce pain, but because of their short duration of action, repeated administration is required. Given frequently, these agents are toxic to the cornea1 epithelium, inhibiting re-epithelialization; prolonged use may cause nonhealing epithelial defects, keratitis, anterior segment inflammation, and visual 10ss.~-’ The toxicity of topical anesthetics also limits their use for other causes of cornea1 pain, such as traumatic abrasions and recurrent erosions. Clearly, there is a need for a long-acting, nontoxic cornea1 anesthetic, not only after photorefractive keratectomy but also for a wide range of injuries to the cornea that cause pain. Tetrodotoxin is a nonprotein neurotoxin found in several vertebrate species, including puffer fish, porcupine fish, goby fish, newts, frogs, and the blue-ringed octopus.’ Since 2500 BCE, human poisonings have been documented after consumption of puffer fish (“fugu”), a delicacy in Japan.‘*‘@ Symptoms of poisonINC. ALL
RIGHTS
RESERVED.
481
ing occur within 10 to 45 minutes of consumption and consist of tingling around the mouth, numbness of the tongue, nausea, vomiting, and paralysis. Death, if it occurs, usually follows consumption by 6 to 24 hours.” Tetrodotoxin toxicity results from its high affinity binding and blockade of neuronal sodium channels that leads to the blockade of action potentials resulting in paralysis, and sensory and autonomic abnormalities.” Because it blocks axonal sodium channels, tetrodotoxin can function as a local anesthetic. This explains the early numbness of oral tissues reported after ingestion. Tetrodotoxin differs from the commonly used synthetic local anesthetics in that it prevents influx of sodium without affecting the efflux of potassium.“$‘4 Ogura and Mori” determined that tetrodotoxin is an effective topical cornea1 anesthetic in rabbits and guinea pigs, producing more than 100 minutes of anesthesia. I5 Unfortunately, these authors noted sig nificant systemic toxicity at effective doses of tetrodotoxin and thus recommended against its usage. A subsequent report by Bynke and associates16 demonstrated up to 4 hours of cornea1 anesthesia after intravitreal injection of tetrodotoxin in rabbits. No tetrodotoxin toxicity was noted. It remains unclear whether tetrodotoxin is sufficiently nontoxic in effective anesthetic doses to be used clinically. However, because there is a need for a longer-acting analgesic for the cornea, we have studied anesthetic duration and toxicity of tetrodotoxin in a rabbit model.
METHODS THE
STUDY
WAS
PERFORMED
IN
ACCORDANCE
WITH
provisions of the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and was approved by the University of California, San Francisco, Committee on Animal Research. Tetrodotoxin in a pH 4.3 sodium citrate vehicle (Sigma Chemical Company, St Louis, Missouri) was obtained and formulated into three concentrations: 10 mM, 1 mM, and 0.1 mM. Twenty-four New Zealand white rabbits were divided into four experimental groups, each consisting of six rabbits. Each group of six rabbits received a 40-p,l aliquot of either tetrodotoxin ( 10 mM, 1 mM, or 0.1 mM) or proparacaine 0.5% (Ophthetic; Allergan, Irvine, California) into 482
AMERICAN
JOURNAL
the inferior conjunctival cul-de-sac of the right eye, with the left eye of each rabbit receiving 40 ~1 of a 60-mM, pH 4.3 sodium citrate vehicle as a control. The experimenters (D.M.S., M.R.J.) who carried out the cornea1 sensation testing were masked to the contents of each aliquot. Cornea1 sensation was tested with a 4-O silk suture mounted upon a wooden cotton tip applicator such that the suture extended 5 mm beyond the wooden end of the applicator. The cornea was stimulated centrally three times with the suture to produce grossly visible indentation of the cornea as the endpoint, similar to a previous rabbit model of cornea1 anesthesia reported by Maurice and Singh.17 Care was taken not to stimulate the eyelashes. The rabbit’s response was graded in the following fashion: no blink = 1, partial blink without full eyelid closure = 2, and full blink = 3. Thus a score of 3 indicates full responsiveness and a score of 1 indicates full anesthesia. The highest anesthesia score of the three tests was recorded for each time point. The data are presented as the mean score of six rabbits. Statistical analysis was by the nonparametric Wilcoxon test with a statistical significance of P 5 .05. Cornea1 sensation was tested before administration of drugs and again at 1 minute, 1 hour, 4 hours, and 8 hours in one experiment or before administration and again at 1 minute, 1 hour, 3 hours, and 5 hours in another experiment. Slit-lamp biomicroscopy with a portable slit lamp was performed with and without fluorescein stain from impregnated strips moistened with balance saline solution at 12 and 24 hours after topical administration. Pachymetry was performed before administration of drugs and again at 24 hours after. The rabbits were observed for changes in feeding habits, movement, respiration, and alertness during the first 24 hours by the experimenters and daily for the subsequent week by animal care personnel.
RESULTS TO ASSESS
WHETHER
TETRODOTOXIN
ADMINlSTRATION
caused clinical alterations in the cornea, we examined all animals with a slit lamp after fluorescein staining. Despite the acidic vehicle, tetrodotoxin administration did not cause any apparent ocular irritation after administration. There was no obvious discomfort in OF OPHTHALMOLOGY
APRIL
1998
Corneal Mean Drug Treatment
Thickness
0 hrs
PROP PROP-V 1mMWX 1 mM TTX-V 10 mM TTX 10 mM TTX-V
0.38 0.39 0.39 0.36 0.36 0.36
'2 + r 3k
lmmi
k SD P Value*
24 hrs
0.03 0.04 0.04 0.01 0.03 0.01
0.36 0.38 0.39 0.39 0.37 0.38
22 0.04 i 0.03 + 0.04 t 0.04 + 0.01 2 0.04
39 39 >.20 -
any of the rabbits evidenced by prolonged eye closure or repetitive blinking. No ocular injection was noted during the 24.hour observation period. At 3 hours, most of the rabbits had a mild central punctate epithelial keratopathy in the area of cornea1 sensation testing, but there was no difference between rabbits that had received proparacaine and those that had received tetrodotoxin. By 24 hours, all signs of epithelial damage had disappeared by slit-lamp examination and fluorescein staining. To evaluate whether endothelial function was significantly affected by tetrodotoxin administration, pachymetry readings before and 24 hours after tetrodotoxin administration were determined on animals treated with 10 mM and 1 mM tetrodotoxin. Pachymetry readings on rabbit eyes that received the highest doses of tetrodotoxin showed no evidence of cornea1 thickening during the 24hour observation period (Table 1). The extent and duration of anesthesia after topical tetrodotoxin varied as a function of dose (Figure 1, Table 2). At 0.1 mM tetrodotoxin, only partial anesthesia was produced in two of six rabbits. At 1 mM tetrodotoxin, anesthesia initially was produced in six of six animals, but the effect was generally short-lived. At 1 minute after administration of tetrodotoxin, the mean anesthesia score was 1.17 (SD = 0.41). At 1 hour, the score was 1.50 (SD = 0.84), VOL.
125,
No.
4
EXPERIMENTAL
STUDY
and by 4 hours, the mean score had increased to 2.83 (SD = 0.41). At 6 h ours, the tetrodotoxin vehicletreated eyes had a better anesthesia score than the l-mM tetrodotoxin-treated eyes did. However, neither of these scores was significantly different from each other or from the anesthesia score of pretreated eyes (score of 3.00; SD = 0). At 10 mM, tetrodotoxin produced a more reproducible and longer-lasting anesthesia. At 1 minute after administration, all rabbit corneas were anesthetic, with a mean anesthesia score of 1 .OO (SD = 0). At 4 hours, anesthesia was still present, with a mean score of 1.17 (SD = 0.41). As late as 8 hours, four of six rabbits showed some residual anesthesia, with a mean score of 2.00 (SD = 0.89). This was significantly different (P = .0325) from the mean score of 3.00 (SD = 0) obtained with vehicle alone at 8 hours (Table 2). We compared the anesthetic duration of 10 mM tetrodotoxin with that of 1 mM tetrodotoxin and proparacaine (Figure 2, Table 3). At the IO-mM dose, tetrodotoxin produced significantly longer anesthesia than proparacaine did. Although proparacaine produced anesthesia in six of six rabbits at 1 minute, by 1 hour, the mean score had increased to 2.50 (SD = 0.84), and at 3 hours, all eyes receiving proparacaine had normal sensation. As late as 5 hours, four of six rabbits that had received tetrodotoxin showed some residual anesthesia, with a mean score of 1.83 (SD = 0.98). This was significantly different (I’ = .0325) from the mean score of 3.00 (SD = 0) obtained with proparacaine or vehicle alone at 5 hours (Table 3). Rabbits were observed carefully for any signs of systemic toxicity during the 24hour observation period. No rabbit had any alterations of feeding, movement, respiration, or alertness during this period that suggested a toxic effect of the tetrodotoxin. No rabbit died or was noted to have abnormalities in behavior by the animal care personnel for 7 days subsequent to the tetrodotoxin administration.
DISCUSSION IN THIS
STUDY
WE DETERMINED
THAT
TETRODOTOXIN
IS
an effective and long-lasting topical anesthetic in a rabbit model. At 10 mM tetrodotoxin, cornea1 anesthesia was rapid in onset and remained nearly comOF TETRODOTOXIN
483
... ...
I 0
0
----(,-e-v
,
I
I
0
60
120
I
I
1mM
I
I
I
I 0
I
180 240 300 360 420 480
TIME
lOmM-V
. ..... 0 ........
lmM-V
_.._ (-J. . . .
O.lrnM
I
-O-
I
60
I
120
O.lmM-V I
I
I
180 240 300 360 420 480
TIME
(min)
I
I
(min)
FIGURE 1. Cornea1 blink response was recorded in rabbit eyes treated with 0.1 mM, 1 mM, or 10 mM tetrodotoxin, as indicated. Data graphed are for blink responses recorded before administration of drugs and after 60, 240, and 480 minutes. The results are graphed as the mean blink response + SEM (indicated by the bars) (N = 6). (Left) Eyes treated with 10 mM, 1 mM, or 0.1 mM tetrodotoxin, as indicated. (Right) Fellow eyes of 0.1 mM tetrodotoxin.treated eyes treated with vehicle alone (0.1 m.M-V), fellow eyes of 1 mM tetrodotoxin-treated eyes treated with vehicle alone (1 mM-V), and fellow eyes of 10 mM tetrodotoxin-treated eyes treated with vehicle alone (10 mM-V), as indicated.
lW,kE
1. Duratian
of An&t&Mia
of Different Slink
Drug Treatment
0.1 mM TTX 0.1 mM TTX-V 1mHTTX 1 mM TTX-V 10mMTTX 10 mM TTX-V
0 min
2.83 2.83 2.83 2.83 3.00 3.00
zt t rt i 2 +
0.41 0.41 0.41 0.41 0.00 0.00
2.67 3.00 1.17 2.83 1.00 3.00
1 n-l,”
P (1 mm)*
k i 2 t + +
5314
0.62 0.00 0.41 0.41 0.00 0.00
c.0022 .OOll -
AMERICAN
(mean
60 ml”
P (60 min)-
2.67 f 0.82
.3496 4.0043 .OOll -
3.00 1.50 2.83 1.00 2.83
+ * -c rt t
plete at 4 hours (Table 2). In most rabbits that received 10 mM tetrodotoxin, significantly reduced sensation lasted at least 8 hours (P = .0325, Wilcoxon test; Table 2). A dose-response relationship was demonstrated, with lower doses producing either shorter or no anesthetic effect. At 10 mM, tetrodotoxin was not only effective but was also well tolerated by the rabbits. After topical administration, there was mild punctate keratopathy but no ocular injection or 484
Response
Concentrations
JOURNAL
0.00 0.84 0.41 0.00 0.41
of Tetrdotoxin
+ SD) (N = 6) 240 min
3.00 3.00 2.63 3.00 1.17 3.00
k t k * + +
0.00 0.00 0.41 0.00 0.41 0.00
P (240 min)’
.99 .3496 ,001 1 -
480 min
2.83 3.00 2.83 2.67 2.00 3.00
2 0.41 t 0.00 k 0.41 + 0.52 * 0.89 z!z 0.00
P (480 min)*
3496 .2944 .0325 -
apparent rabbit discomfort. No rabbit showed any sign of systemic toxicity for up to 1 week after topical administration. The local anesthetic properties of crude tetrodotoxin extracts have been long recognized in Japan.la~Zo After purification of tetrodotoxin in the 195Os, Kao and Fuhrman’” established that tetrodotoxin was 160,000 times more potent than cocaine as an anesthetic. Although conduction block in crayfish OF OPHTHALMOLOGY
APRIL
I 998
--iF-
--c)--
10m
lOmM-V
.......0 ....... lmM-V I 0
.J I 0
ho
40
1;o
TIME
260
0
3AO
-_._ 0 .___ I 60
I 120
TIME
(min)
I 180
I 240
PROP-V 1 300
(min)
FIGURE 2. Cornea1 blink response was recorded in rabbit eyes treated with 1 mM or 10 mM tetrodotoxin or with proparacaine (PROP), as indicated. Data are graphed for blink responses recorded before administration of drugs and after 60, 180, and 300 minutes. The results are graphed in the figure as the mean blink response it SEM (indicated by the bars) (N = 6). (Left) Eyes treated with 10 mM or 1 mM tetrodotoxin or with proparacaine (PROP), as indicated. (Right) Fellow eyes of 10 mM tetrodotoxin-treated eyes treated with vehicle alone (10 r&l-V), fellow eyes of 1 mM eyes treated with tetrodotoxin-treated eyes treated with vehicle alone (1 n&i-V), an d fellow eyes of proparacaine-treated vehicle alone (PROP-V), as indicated.
nerve fibers was produced with lo-l3 M tetrodotoxin, topical cornea1 application of tetrodotoxin in rabbits required much higher concentrations for anesthesia.16 Cornea1 application of 0.9 PM tetrodotoxin produced a mean anesthesia duration of 34.8 minutes in rabbits, which increased to “>lOO” minutes with 30 p,M tetrodotoxin.16 The reason for the marked difference in sensitivity to tetrodotoxin by isolated axons and the in vivo cornea is unclear but may be related to both the anatomic distribution of cornea1 nerves and the polarity of tetrodotoxin. The unmyelinated nerve fibers that innervate the cornea emerge from the stroma to supply all but the two most superficial layers of the epithelium.2’ To anesthetize the cornea, tetrodotoxin must penetrate these superficial epithelial layers to bind to sodium channels of the sensory axons. Tetrodotoxin is a zwitterion with a pKa = 8.5.“1’~ Its activity is significantly enhanced in its cationic form at the acidic pH of its citrate vehicle. I6 At this pH, tetrodotoxin is quite polar and would be expected to have limited cornea1 penetration. ” Therefore, the high concentrations of tetrodotoxin required to produce cornea1 VOL.
125,
No.
4
EXPERIMENTAL
STUDY
anesthesia in this study may reflect poor penetration of the superficial cornea1 epithelium by highly polar tetrodotoxin molecules. Despite the longer-acting anesthesia produced by tetrodotoxin administration in a rabbit model, Ogura and associates” recommended against pursuing clinical usage because of potential tetrodotoxin toxicity. At 30 PM, a dose of topical tetrodotoxin produced systemic toxicity in guinea pigs; animals were reported as “near to death.” Interestingly, we found no evidence of either ocular or systemic toxicity of a dose of topical 10 mM tetrodotoxin in rabbits, a 333-fold increase in the toxic concentration reported for guinea pigs. This discrepancy in the toxicity is unclear but could be related to differences in the purity of the tetrodotoxin, its vehicle, or different susceptibility to toxic effects by rabbits and guinea pigs. Despite the absence of any apparent systemic toxicity in the present investigation, the concentration of tetrodotoxin used is potentially toxic, even in humans. The dose lethality curve for tetrodotoxin is very steep. In mice, the minimum intraperitoneal OF TETRODOTOXIN
485
Blink Response Drug Treatment
t&rodotoxln-tre
0 ml”
yes;
1 min
TTX-V
P (1 mm)’
60 min
i- SD) (N = 6) 160 min
P (160 min)’
300 min
P (300 min)’
=
lethal dose is 8 pg/kg and the LD,, is 12 w&g.2” The lethal dose in mice is much higher for oral administration, 332 kg/kg.‘” Lethal doses have been determined experimentally for multiple animal species,” but the dose for human lethality has been only imprecisely estimated. Cornish’i reported a case of tetrodotoxin intoxication and estimated the lethal oral dose to be about 10 to 18 kg/kg based on the amount of fish tissue consutned. For a 70-kg person, the lethal dose would therefore be 0.7 to 1.26 mg. A 40,~1 aliquot of 10 mM tetrodotoxin delivers 127 kg of tetrodotoxin to the rabbit’s eye. This is probably within 1 log unit of the lethal oral dose. The lethal dose after absorption through the nasal mucosa may be lower than the lethal oral dose, indicating that topical administration to the cornea at 10 mM may be even more dangerous. Systemic absorption after topical administration has not been determined, It is possible that in a post-photorefractive keratectomy, de-epithelialized cornea, reductions in the effective anesthetic dose of tetrodotoxin might be achieved, thus increasing the safety margin of this potential anesthetic agent. Additionally, the human eye will hold less than 20 ~1, so further reduction in the effective anesthetic dose may be achieved by administering a smaller drop volume than that used in this study. These issues need to be addressed in future studies before tetrodotoxin can safely be applied to humans. In summary, in the current study, we have determined that tetrodotoxin is a very long-acting topical anesthetic in the rabbit cornea. Tetrodotoxin at a
486
(mean P (60 min)’
AMERICAN
JOURNAL
concentration of 10 mM showed no evidence of significant ocular or systemic toxicity. Despite the absence of observed toxicity, tetrodotoxin is a highly toxic substance, with no available antidote, and further study of systemic absorption after topical administration is warranted. We are currently addressing the effects of this anesthetic on the deepithelialized cornea as well as measuring systemic absorption after topical administration with a cell culture bioassay. 24 If tetrodotoxin is effective as a long-term anesthetic when applied at safe doses to a de-epithelialized cornea, it may provide an excellent means to control postoperative pain of photorefractive keratectomy.
REFERENCES 1. Cherry PM. The treatment of pain following excimer laser photorefractive keratectomy: additive effect of local anesthetic drops, topical diclofenac, and bandage soft contact. J Cataract Refract Surg 1996;27(5,suppl):S477-S480. 2. Verma S, Marshall J. Control of pain after photorefractive keratectomy. J Refract Surg 1996;12:358-364. 3. Tutcon MK, Cherry PM, Raj PS, Fsadni MG. Efficacy and safety of topical diclofenac in reducing ocular pain after excimer photorefractive keratectomy. J Cataract Refract Surg 1996;22:536-541. 4. Smith RB, Everett WG. Physiology and pharmacology of local anesthetic agents. Inc Ophthalmol Clin 1973;13: 35-60. 5. Behrendt T. Experimental study of cornea1 lesions: produced by topical anesthesia. Am J Ophthalmol 1956;41:99-105. 6. Rosenwasser GOD. Complications of topical ocular anesthetics. Intl Ophthalmol Clin 1989;29:153-158.
OF OPHTHALMOLOGY
APRIL
1998
7. Rosenwasser GOD, Holland S, Pflugfelder SC, et al. Topical anesthetic abuse. Ophthalmology 1990;97:967-972. 8. Bower DJ, Hart, RJ, Matthews PA, Howden ME. Nonprotein neurotoxins. Clin Toxic01 1981;18:813-863. 9. Mills AR, Passmore R. Pelagic paralysis. Lancet 1988; 1:161-164. 10. Kao CY. Tetrodotoxin, saxitoxin and their significance in the study of excitation phenomena. Pharm Review 1966;18: 997-1049. II. Tetrodotoxin poisoning associated with eating puffer fish transported from Japan-California. MMWR Morb Mortal Wkly Rep 1996;45:389-391. 12. Lange WR. Puffer fish poisoning. Am Fam Physician 1990; 42:1029-1033. 13. Mosher HS, Fuhrman A, Buchwald HD, Fischer HG. Tarichatoxin-tetrodotoxin: a potent neurotoxin. Science 1964;144:1100-1110. 14. Blankenship JE. Tetrodotoxin: from poison to powerful tool. Perspect in Biol and Med 1976;19:509-526. 15. Ogura Y, Mori Y. Mechanism of local anesthetic action of crystalline tetrodotoxin and its derivatives. Eur J Pharmacol 1968:3:58-67. 16. Bynke G, Hakanson R, Sundler F. Is substance P necessary for cornea1 nociception? Eur J Pharmacol 1984;101:253-258.
17. Maurice DM, Singh T. The absence of cornea1 toxicity with low-level topical anesthesia. Am J Ophthalmol 1985;99: 691-696. 18. Ogura Y, Mori Y, Watanabe Y. Local anesthetic activity of crystalline tetrodotoxin and its related compounds. Folia Pharmacol 1966:Sect 20:62. 19. Ogura Y, Hamada J, Tsukada 0. Action of crystalline tetrodotoxin on nicotine and serotonin. Folia Pharmacol 1960;Sect 35:56. 20. Kao CY, Fuhrman FA. Pharmacological studies on tarichatoxin, a potent neurotoxin. J Pharmacol 1963;140:31-40. 21. Spencer WH. Ophthalmic pathology, 3rd ed. Philadelphia: WB Saunders, 1985:235. 22. Unger WG, Cole DF, Bass MS. Prostaglandin and neurogenically mediated ocular response to laser irradiation of the rabbit iris. Exp Eye Res 1977;25:209-220. 23. Cornish JN. Susceptibility of man to puffer fish toxin. Austral 1973;2:48. 24. Hamasaki K, Kogure K, Ohwada K. A biological method for the quantitative measurement of tetrodotoxin (TTX): tissue culture bioassay in combination with a water-soluble tetrazolium salt. Toxicon 1996:34:490-495.
Authors Interactive@ We encourage questions and comments regarding this article via the Internet on Authors Interactive@ at http://www.ajo.com/ Questions, comments, and author responses are posted.
VOL. 125,
No.
4
EXPERIMENTAL
STUDY
OF TETRODOTOXIN
487
PURPOSE:
Accepted for publication July 24, 1997. From the Department of Ophthalmology (Drs Schwartz, K. Duncan. I. Duncan, and Jones) and the-Departments of Neurology and Physiology (Dr Fields), Universitv of California San Francisco Medical Center. San Francisco, California. Supported in part by an unrestricted grant and core grant EY02162 from Research to Prevent Blindness, Inc, New York, New York. Drs Schwartz and Fields have filed a patent for use of tetrodotoxin. Reprint requests to Daniel M. Schwartz, MD, Department of Ophthalmology, UCSF, 10 Kickham St, Box 0730, San Francisco, CA 941430730; fax: (415) 476-0336; e-mail: [email protected]
0002-9394/98/$19.00
0
1998
BY ELSEVIER SCIENCE
this effect was largely gone by 1 hour and complete0 ly gone by 3 hours. l CONCLUSIONS: Tetrodotoxin is a long-acting topical anesthetic in the rabbit cornea. Although additional toxicity studies are required, tetrodoe toxin may provide an effective, long-lasting topical anesthetic for use in pain control after cornea1 procedures such as photorefractive keratectomy. (Am J Ophthalmol 1998;125:481-487. 0 1998 by Elsevier Science Inc. All rights reserved.)
P
OSTOPERATIVE
PAIN
CAN
BE A DISABLING
COM-
plication of photorefractive keratectomy. Current management with bandage contact lens, nonsteroidal anti-inflammatory agents, and oral analgesics mitigates, but does not eliminate, the discomfort in most patients. 1-3Topical anesthetics have been used to reduce pain, but because of their short duration of action, repeated administration is required. Given frequently, these agents are toxic to the cornea1 epithelium, inhibiting re-epithelialization; prolonged use may cause nonhealing epithelial defects, keratitis, anterior segment inflammation, and visual 10ss.~-’ The toxicity of topical anesthetics also limits their use for other causes of cornea1 pain, such as traumatic abrasions and recurrent erosions. Clearly, there is a need for a long-acting, nontoxic cornea1 anesthetic, not only after photorefractive keratectomy but also for a wide range of injuries to the cornea that cause pain. Tetrodotoxin is a nonprotein neurotoxin found in several vertebrate species, including puffer fish, porcupine fish, goby fish, newts, frogs, and the blue-ringed octopus.’ Since 2500 BCE, human poisonings have been documented after consumption of puffer fish (“fugu”), a delicacy in Japan.‘*‘@ Symptoms of poisonINC. ALL
RIGHTS
RESERVED.
481
ing occur within 10 to 45 minutes of consumption and consist of tingling around the mouth, numbness of the tongue, nausea, vomiting, and paralysis. Death, if it occurs, usually follows consumption by 6 to 24 hours.” Tetrodotoxin toxicity results from its high affinity binding and blockade of neuronal sodium channels that leads to the blockade of action potentials resulting in paralysis, and sensory and autonomic abnormalities.” Because it blocks axonal sodium channels, tetrodotoxin can function as a local anesthetic. This explains the early numbness of oral tissues reported after ingestion. Tetrodotoxin differs from the commonly used synthetic local anesthetics in that it prevents influx of sodium without affecting the efflux of potassium.“$‘4 Ogura and Mori” determined that tetrodotoxin is an effective topical cornea1 anesthetic in rabbits and guinea pigs, producing more than 100 minutes of anesthesia. I5 Unfortunately, these authors noted sig nificant systemic toxicity at effective doses of tetrodotoxin and thus recommended against its usage. A subsequent report by Bynke and associates16 demonstrated up to 4 hours of cornea1 anesthesia after intravitreal injection of tetrodotoxin in rabbits. No tetrodotoxin toxicity was noted. It remains unclear whether tetrodotoxin is sufficiently nontoxic in effective anesthetic doses to be used clinically. However, because there is a need for a longer-acting analgesic for the cornea, we have studied anesthetic duration and toxicity of tetrodotoxin in a rabbit model.
METHODS THE
STUDY
WAS
PERFORMED
IN
ACCORDANCE
WITH
provisions of the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and was approved by the University of California, San Francisco, Committee on Animal Research. Tetrodotoxin in a pH 4.3 sodium citrate vehicle (Sigma Chemical Company, St Louis, Missouri) was obtained and formulated into three concentrations: 10 mM, 1 mM, and 0.1 mM. Twenty-four New Zealand white rabbits were divided into four experimental groups, each consisting of six rabbits. Each group of six rabbits received a 40-p,l aliquot of either tetrodotoxin ( 10 mM, 1 mM, or 0.1 mM) or proparacaine 0.5% (Ophthetic; Allergan, Irvine, California) into 482
AMERICAN
JOURNAL
the inferior conjunctival cul-de-sac of the right eye, with the left eye of each rabbit receiving 40 ~1 of a 60-mM, pH 4.3 sodium citrate vehicle as a control. The experimenters (D.M.S., M.R.J.) who carried out the cornea1 sensation testing were masked to the contents of each aliquot. Cornea1 sensation was tested with a 4-O silk suture mounted upon a wooden cotton tip applicator such that the suture extended 5 mm beyond the wooden end of the applicator. The cornea was stimulated centrally three times with the suture to produce grossly visible indentation of the cornea as the endpoint, similar to a previous rabbit model of cornea1 anesthesia reported by Maurice and Singh.17 Care was taken not to stimulate the eyelashes. The rabbit’s response was graded in the following fashion: no blink = 1, partial blink without full eyelid closure = 2, and full blink = 3. Thus a score of 3 indicates full responsiveness and a score of 1 indicates full anesthesia. The highest anesthesia score of the three tests was recorded for each time point. The data are presented as the mean score of six rabbits. Statistical analysis was by the nonparametric Wilcoxon test with a statistical significance of P 5 .05. Cornea1 sensation was tested before administration of drugs and again at 1 minute, 1 hour, 4 hours, and 8 hours in one experiment or before administration and again at 1 minute, 1 hour, 3 hours, and 5 hours in another experiment. Slit-lamp biomicroscopy with a portable slit lamp was performed with and without fluorescein stain from impregnated strips moistened with balance saline solution at 12 and 24 hours after topical administration. Pachymetry was performed before administration of drugs and again at 24 hours after. The rabbits were observed for changes in feeding habits, movement, respiration, and alertness during the first 24 hours by the experimenters and daily for the subsequent week by animal care personnel.
RESULTS TO ASSESS
WHETHER
TETRODOTOXIN
ADMINlSTRATION
caused clinical alterations in the cornea, we examined all animals with a slit lamp after fluorescein staining. Despite the acidic vehicle, tetrodotoxin administration did not cause any apparent ocular irritation after administration. There was no obvious discomfort in OF OPHTHALMOLOGY
APRIL
1998
Corneal Mean Drug Treatment
Thickness
0 hrs
PROP PROP-V 1mMWX 1 mM TTX-V 10 mM TTX 10 mM TTX-V
0.38 0.39 0.39 0.36 0.36 0.36
'2 + r 3k
lmmi
k SD P Value*
24 hrs
0.03 0.04 0.04 0.01 0.03 0.01
0.36 0.38 0.39 0.39 0.37 0.38
22 0.04 i 0.03 + 0.04 t 0.04 + 0.01 2 0.04
39 39 >.20 -
any of the rabbits evidenced by prolonged eye closure or repetitive blinking. No ocular injection was noted during the 24.hour observation period. At 3 hours, most of the rabbits had a mild central punctate epithelial keratopathy in the area of cornea1 sensation testing, but there was no difference between rabbits that had received proparacaine and those that had received tetrodotoxin. By 24 hours, all signs of epithelial damage had disappeared by slit-lamp examination and fluorescein staining. To evaluate whether endothelial function was significantly affected by tetrodotoxin administration, pachymetry readings before and 24 hours after tetrodotoxin administration were determined on animals treated with 10 mM and 1 mM tetrodotoxin. Pachymetry readings on rabbit eyes that received the highest doses of tetrodotoxin showed no evidence of cornea1 thickening during the 24hour observation period (Table 1). The extent and duration of anesthesia after topical tetrodotoxin varied as a function of dose (Figure 1, Table 2). At 0.1 mM tetrodotoxin, only partial anesthesia was produced in two of six rabbits. At 1 mM tetrodotoxin, anesthesia initially was produced in six of six animals, but the effect was generally short-lived. At 1 minute after administration of tetrodotoxin, the mean anesthesia score was 1.17 (SD = 0.41). At 1 hour, the score was 1.50 (SD = 0.84), VOL.
125,
No.
4
EXPERIMENTAL
STUDY
and by 4 hours, the mean score had increased to 2.83 (SD = 0.41). At 6 h ours, the tetrodotoxin vehicletreated eyes had a better anesthesia score than the l-mM tetrodotoxin-treated eyes did. However, neither of these scores was significantly different from each other or from the anesthesia score of pretreated eyes (score of 3.00; SD = 0). At 10 mM, tetrodotoxin produced a more reproducible and longer-lasting anesthesia. At 1 minute after administration, all rabbit corneas were anesthetic, with a mean anesthesia score of 1 .OO (SD = 0). At 4 hours, anesthesia was still present, with a mean score of 1.17 (SD = 0.41). As late as 8 hours, four of six rabbits showed some residual anesthesia, with a mean score of 2.00 (SD = 0.89). This was significantly different (P = .0325) from the mean score of 3.00 (SD = 0) obtained with vehicle alone at 8 hours (Table 2). We compared the anesthetic duration of 10 mM tetrodotoxin with that of 1 mM tetrodotoxin and proparacaine (Figure 2, Table 3). At the IO-mM dose, tetrodotoxin produced significantly longer anesthesia than proparacaine did. Although proparacaine produced anesthesia in six of six rabbits at 1 minute, by 1 hour, the mean score had increased to 2.50 (SD = 0.84), and at 3 hours, all eyes receiving proparacaine had normal sensation. As late as 5 hours, four of six rabbits that had received tetrodotoxin showed some residual anesthesia, with a mean score of 1.83 (SD = 0.98). This was significantly different (I’ = .0325) from the mean score of 3.00 (SD = 0) obtained with proparacaine or vehicle alone at 5 hours (Table 3). Rabbits were observed carefully for any signs of systemic toxicity during the 24hour observation period. No rabbit had any alterations of feeding, movement, respiration, or alertness during this period that suggested a toxic effect of the tetrodotoxin. No rabbit died or was noted to have abnormalities in behavior by the animal care personnel for 7 days subsequent to the tetrodotoxin administration.
DISCUSSION IN THIS
STUDY
WE DETERMINED
THAT
TETRODOTOXIN
IS
an effective and long-lasting topical anesthetic in a rabbit model. At 10 mM tetrodotoxin, cornea1 anesthesia was rapid in onset and remained nearly comOF TETRODOTOXIN
483
... ...
I 0
0
----(,-e-v
,
I
I
0
60
120
I
I
1mM
I
I
I
I 0
I
180 240 300 360 420 480
TIME
lOmM-V
. ..... 0 ........
lmM-V
_.._ (-J. . . .
O.lrnM
I
-O-
I
60
I
120
O.lmM-V I
I
I
180 240 300 360 420 480
TIME
(min)
I
I
(min)
FIGURE 1. Cornea1 blink response was recorded in rabbit eyes treated with 0.1 mM, 1 mM, or 10 mM tetrodotoxin, as indicated. Data graphed are for blink responses recorded before administration of drugs and after 60, 240, and 480 minutes. The results are graphed as the mean blink response + SEM (indicated by the bars) (N = 6). (Left) Eyes treated with 10 mM, 1 mM, or 0.1 mM tetrodotoxin, as indicated. (Right) Fellow eyes of 0.1 mM tetrodotoxin.treated eyes treated with vehicle alone (0.1 m.M-V), fellow eyes of 1 mM tetrodotoxin-treated eyes treated with vehicle alone (1 mM-V), and fellow eyes of 10 mM tetrodotoxin-treated eyes treated with vehicle alone (10 mM-V), as indicated.
lW,kE
1. Duratian
of An&t&Mia
of Different Slink
Drug Treatment
0.1 mM TTX 0.1 mM TTX-V 1mHTTX 1 mM TTX-V 10mMTTX 10 mM TTX-V
0 min
2.83 2.83 2.83 2.83 3.00 3.00
zt t rt i 2 +
0.41 0.41 0.41 0.41 0.00 0.00
2.67 3.00 1.17 2.83 1.00 3.00
1 n-l,”
P (1 mm)*
k i 2 t + +
5314
0.62 0.00 0.41 0.41 0.00 0.00
c.0022 .OOll -
AMERICAN
(mean
60 ml”
P (60 min)-
2.67 f 0.82
.3496 4.0043 .OOll -
3.00 1.50 2.83 1.00 2.83
+ * -c rt t
plete at 4 hours (Table 2). In most rabbits that received 10 mM tetrodotoxin, significantly reduced sensation lasted at least 8 hours (P = .0325, Wilcoxon test; Table 2). A dose-response relationship was demonstrated, with lower doses producing either shorter or no anesthetic effect. At 10 mM, tetrodotoxin was not only effective but was also well tolerated by the rabbits. After topical administration, there was mild punctate keratopathy but no ocular injection or 484
Response
Concentrations
JOURNAL
0.00 0.84 0.41 0.00 0.41
of Tetrdotoxin
+ SD) (N = 6) 240 min
3.00 3.00 2.63 3.00 1.17 3.00
k t k * + +
0.00 0.00 0.41 0.00 0.41 0.00
P (240 min)’
.99 .3496 ,001 1 -
480 min
2.83 3.00 2.83 2.67 2.00 3.00
2 0.41 t 0.00 k 0.41 + 0.52 * 0.89 z!z 0.00
P (480 min)*
3496 .2944 .0325 -
apparent rabbit discomfort. No rabbit showed any sign of systemic toxicity for up to 1 week after topical administration. The local anesthetic properties of crude tetrodotoxin extracts have been long recognized in Japan.la~Zo After purification of tetrodotoxin in the 195Os, Kao and Fuhrman’” established that tetrodotoxin was 160,000 times more potent than cocaine as an anesthetic. Although conduction block in crayfish OF OPHTHALMOLOGY
APRIL
I 998
--iF-
--c)--
10m
lOmM-V
.......0 ....... lmM-V I 0
.J I 0
ho
40
1;o
TIME
260
0
3AO
-_._ 0 .___ I 60
I 120
TIME
(min)
I 180
I 240
PROP-V 1 300
(min)
FIGURE 2. Cornea1 blink response was recorded in rabbit eyes treated with 1 mM or 10 mM tetrodotoxin or with proparacaine (PROP), as indicated. Data are graphed for blink responses recorded before administration of drugs and after 60, 180, and 300 minutes. The results are graphed in the figure as the mean blink response it SEM (indicated by the bars) (N = 6). (Left) Eyes treated with 10 mM or 1 mM tetrodotoxin or with proparacaine (PROP), as indicated. (Right) Fellow eyes of 10 mM tetrodotoxin-treated eyes treated with vehicle alone (10 r&l-V), fellow eyes of 1 mM eyes treated with tetrodotoxin-treated eyes treated with vehicle alone (1 n&i-V), an d fellow eyes of proparacaine-treated vehicle alone (PROP-V), as indicated.
nerve fibers was produced with lo-l3 M tetrodotoxin, topical cornea1 application of tetrodotoxin in rabbits required much higher concentrations for anesthesia.16 Cornea1 application of 0.9 PM tetrodotoxin produced a mean anesthesia duration of 34.8 minutes in rabbits, which increased to “>lOO” minutes with 30 p,M tetrodotoxin.16 The reason for the marked difference in sensitivity to tetrodotoxin by isolated axons and the in vivo cornea is unclear but may be related to both the anatomic distribution of cornea1 nerves and the polarity of tetrodotoxin. The unmyelinated nerve fibers that innervate the cornea emerge from the stroma to supply all but the two most superficial layers of the epithelium.2’ To anesthetize the cornea, tetrodotoxin must penetrate these superficial epithelial layers to bind to sodium channels of the sensory axons. Tetrodotoxin is a zwitterion with a pKa = 8.5.“1’~ Its activity is significantly enhanced in its cationic form at the acidic pH of its citrate vehicle. I6 At this pH, tetrodotoxin is quite polar and would be expected to have limited cornea1 penetration. ” Therefore, the high concentrations of tetrodotoxin required to produce cornea1 VOL.
125,
No.
4
EXPERIMENTAL
STUDY
anesthesia in this study may reflect poor penetration of the superficial cornea1 epithelium by highly polar tetrodotoxin molecules. Despite the longer-acting anesthesia produced by tetrodotoxin administration in a rabbit model, Ogura and associates” recommended against pursuing clinical usage because of potential tetrodotoxin toxicity. At 30 PM, a dose of topical tetrodotoxin produced systemic toxicity in guinea pigs; animals were reported as “near to death.” Interestingly, we found no evidence of either ocular or systemic toxicity of a dose of topical 10 mM tetrodotoxin in rabbits, a 333-fold increase in the toxic concentration reported for guinea pigs. This discrepancy in the toxicity is unclear but could be related to differences in the purity of the tetrodotoxin, its vehicle, or different susceptibility to toxic effects by rabbits and guinea pigs. Despite the absence of any apparent systemic toxicity in the present investigation, the concentration of tetrodotoxin used is potentially toxic, even in humans. The dose lethality curve for tetrodotoxin is very steep. In mice, the minimum intraperitoneal OF TETRODOTOXIN
485
Blink Response Drug Treatment
t&rodotoxln-tre
0 ml”
yes;
1 min
TTX-V
P (1 mm)’
60 min
i- SD) (N = 6) 160 min
P (160 min)’
300 min
P (300 min)’
=
lethal dose is 8 pg/kg and the LD,, is 12 w&g.2” The lethal dose in mice is much higher for oral administration, 332 kg/kg.‘” Lethal doses have been determined experimentally for multiple animal species,” but the dose for human lethality has been only imprecisely estimated. Cornish’i reported a case of tetrodotoxin intoxication and estimated the lethal oral dose to be about 10 to 18 kg/kg based on the amount of fish tissue consutned. For a 70-kg person, the lethal dose would therefore be 0.7 to 1.26 mg. A 40,~1 aliquot of 10 mM tetrodotoxin delivers 127 kg of tetrodotoxin to the rabbit’s eye. This is probably within 1 log unit of the lethal oral dose. The lethal dose after absorption through the nasal mucosa may be lower than the lethal oral dose, indicating that topical administration to the cornea at 10 mM may be even more dangerous. Systemic absorption after topical administration has not been determined, It is possible that in a post-photorefractive keratectomy, de-epithelialized cornea, reductions in the effective anesthetic dose of tetrodotoxin might be achieved, thus increasing the safety margin of this potential anesthetic agent. Additionally, the human eye will hold less than 20 ~1, so further reduction in the effective anesthetic dose may be achieved by administering a smaller drop volume than that used in this study. These issues need to be addressed in future studies before tetrodotoxin can safely be applied to humans. In summary, in the current study, we have determined that tetrodotoxin is a very long-acting topical anesthetic in the rabbit cornea. Tetrodotoxin at a
486
(mean P (60 min)’
AMERICAN
JOURNAL
concentration of 10 mM showed no evidence of significant ocular or systemic toxicity. Despite the absence of observed toxicity, tetrodotoxin is a highly toxic substance, with no available antidote, and further study of systemic absorption after topical administration is warranted. We are currently addressing the effects of this anesthetic on the deepithelialized cornea as well as measuring systemic absorption after topical administration with a cell culture bioassay. 24 If tetrodotoxin is effective as a long-term anesthetic when applied at safe doses to a de-epithelialized cornea, it may provide an excellent means to control postoperative pain of photorefractive keratectomy.
REFERENCES 1. Cherry PM. The treatment of pain following excimer laser photorefractive keratectomy: additive effect of local anesthetic drops, topical diclofenac, and bandage soft contact. J Cataract Refract Surg 1996;27(5,suppl):S477-S480. 2. Verma S, Marshall J. Control of pain after photorefractive keratectomy. J Refract Surg 1996;12:358-364. 3. Tutcon MK, Cherry PM, Raj PS, Fsadni MG. Efficacy and safety of topical diclofenac in reducing ocular pain after excimer photorefractive keratectomy. J Cataract Refract Surg 1996;22:536-541. 4. Smith RB, Everett WG. Physiology and pharmacology of local anesthetic agents. Inc Ophthalmol Clin 1973;13: 35-60. 5. Behrendt T. Experimental study of cornea1 lesions: produced by topical anesthesia. Am J Ophthalmol 1956;41:99-105. 6. Rosenwasser GOD. Complications of topical ocular anesthetics. Intl Ophthalmol Clin 1989;29:153-158.
OF OPHTHALMOLOGY
APRIL
1998
7. Rosenwasser GOD, Holland S, Pflugfelder SC, et al. Topical anesthetic abuse. Ophthalmology 1990;97:967-972. 8. Bower DJ, Hart, RJ, Matthews PA, Howden ME. Nonprotein neurotoxins. Clin Toxic01 1981;18:813-863. 9. Mills AR, Passmore R. Pelagic paralysis. Lancet 1988; 1:161-164. 10. Kao CY. Tetrodotoxin, saxitoxin and their significance in the study of excitation phenomena. Pharm Review 1966;18: 997-1049. II. Tetrodotoxin poisoning associated with eating puffer fish transported from Japan-California. MMWR Morb Mortal Wkly Rep 1996;45:389-391. 12. Lange WR. Puffer fish poisoning. Am Fam Physician 1990; 42:1029-1033. 13. Mosher HS, Fuhrman A, Buchwald HD, Fischer HG. Tarichatoxin-tetrodotoxin: a potent neurotoxin. Science 1964;144:1100-1110. 14. Blankenship JE. Tetrodotoxin: from poison to powerful tool. Perspect in Biol and Med 1976;19:509-526. 15. Ogura Y, Mori Y. Mechanism of local anesthetic action of crystalline tetrodotoxin and its derivatives. Eur J Pharmacol 1968:3:58-67. 16. Bynke G, Hakanson R, Sundler F. Is substance P necessary for cornea1 nociception? Eur J Pharmacol 1984;101:253-258.
17. Maurice DM, Singh T. The absence of cornea1 toxicity with low-level topical anesthesia. Am J Ophthalmol 1985;99: 691-696. 18. Ogura Y, Mori Y, Watanabe Y. Local anesthetic activity of crystalline tetrodotoxin and its related compounds. Folia Pharmacol 1966:Sect 20:62. 19. Ogura Y, Hamada J, Tsukada 0. Action of crystalline tetrodotoxin on nicotine and serotonin. Folia Pharmacol 1960;Sect 35:56. 20. Kao CY, Fuhrman FA. Pharmacological studies on tarichatoxin, a potent neurotoxin. J Pharmacol 1963;140:31-40. 21. Spencer WH. Ophthalmic pathology, 3rd ed. Philadelphia: WB Saunders, 1985:235. 22. Unger WG, Cole DF, Bass MS. Prostaglandin and neurogenically mediated ocular response to laser irradiation of the rabbit iris. Exp Eye Res 1977;25:209-220. 23. Cornish JN. Susceptibility of man to puffer fish toxin. Austral 1973;2:48. 24. Hamasaki K, Kogure K, Ohwada K. A biological method for the quantitative measurement of tetrodotoxin (TTX): tissue culture bioassay in combination with a water-soluble tetrazolium salt. Toxicon 1996:34:490-495.
Authors Interactive@ We encourage questions and comments regarding this article via the Internet on Authors Interactive@ at http://www.ajo.com/ Questions, comments, and author responses are posted.
VOL. 125,
No.
4
EXPERIMENTAL
STUDY
OF TETRODOTOXIN
487