Ciguatera

Ciguatera is a clinical syndrome caused by eating the flesh of toxic fish caught in tropical reef and island waters. The toxin is believed to originate in a microscopic dinoflagellate alga Gambierdiscus toxicus that grows on reefs (Bagnis et al., 1980). However, other benthic algae have also been implicated. Fish eating the algae become toxic, and the effect is magnified through the food chain so that large predatory fish become the most toxic. The occurrence of toxic fish tends to be localized, but localization is not consistent and toxic fish may occur sporadically anywhere in a reef or island location (Engleberg et al., 1983). More than 400 species have been implicated in ciguatera poisoning (Randall, 1980), but the fish most commonly implicated include amberjack, snapper, grouper, barracuda, goatfish, and reef fish belonging to the Carrangidae (Table 4-2). In the United States, ciguatera occurs principally in Hawaii, Puerto Rico, the Virgin Islands, Guam, and Florida (CDC, 1989). A particularly high incidence was reported from Guam (Haddock, 1989), and a few cases have been reported in other states caused by fish shipped from Florida. Cases are frequently associated with travel to endemic ciguatera areas such as Hawaii and the Virgin Islands, and there is concern that many cases are not recognized by mainland U.S. physicians.

TABLE 4-2

Fish Reported as a Vehicle of Ciguatera to CDC 1978-1987 Ordered by Frequency of Reported Involvement.

The disease affects both gastrointestinal and neurological systems (Bagnis et al., 1979; Morris et al., 1982a). Gastrointestinal symptoms, including diarrhea, nausea, vomiting, and abdominal pain, appear 3-5 hours after ingestion of the fish and are of short duration. Neurological symptoms begin 12-18 hours after consumption and may be moderate to severe; they commonly last for 1-82 days but may persist for several months. In rare cases, symptoms may last for years, with exacerbation associated with fish consumption or possibly alcohol (Halstead, 1967). Symptoms typically include hot-cold inversion (hot coffee tastes cold, ice cream tastes hot); muscular aches; tingling and numbness of lips, tongue, and perioral region; metallic taste; dryness of mouth; anxiety; prostration; dizziness; chills; sweating; dilated eyes, blurred vision, and temporary blindness. Paralysis and death may occur in a few extreme cases. Symptoms may be extremely debilitating, resulting in extended periods of disability. Intravenous mannitol may relieve acute symptoms (Palafox et al., 1988), provided it is given within several hours of consumption, with amitriptyline (Bowman, 1984) or tocainide (Lange et al., 1988) suggested for more chronic manifestations. There is considerable individuality in patient response (Engleberg et al., 1983).

Several toxic compounds have been isolated from ciguatoxic fish and from Gambierdiscus. The principal toxin called "ciguatoxin" is a small lipid-soluble polyether with a molecular weight of 1,112 (Scheuer et al., 1967); this toxin has been purified and its structure determined (Murata et al., 1990). Ciguatoxin (CTX) has a molecular formula of C₆₀H₈₈O₁₉ and is a brevitoxin type polyether, approximately 100 times more potent than terodotoxin. Ciguatoxin opens voltage-dependent sodium channels in cell membranes (Bidard, 1984), and studies with in vitro tissue preparations suggest that the toxin causes a nerve conduction block after initial neural stimulation. In animal models, low doses of ciguatoxin cause mild hypotension and brachycardia. Higher doses give a biphasic response with an initial brachycardia/hypotension followed by tachycardia/hypertension; very high doses produce a phrenic nerve block with respiratory arrest (Gillespie et al., 1986). Another lipid-soluble neurotoxin from ciguateric fish is called "scaritoxin." This toxin has been shown to depress oxidative metabolic processes in rat brain and has a depolarizing action on excitable membranes.

Generally, the pharmacological action is close to that of ciguatoxin, and they may be related compounds (Legrand and Bagnis, 1984). Maitotoxin is a water-soluble toxin that may interfere with or modify calcium movement or calcium conductance in tissues. Other lipid-soluble toxins have been reported, but their structures and pharmacologic roles are not understood (Ragelis, 1984).

The reported incidence of ciguatera as indicated by CDC data is on the order of 15-20 outbreaks per year, involving 50 to 100 cases. Cases reported to the CDC occur almost exclusively in Hawaii, Puerto Rico, the Virgin Islands, and Florida (Table 4-3). However, these numbers appear to reflect significant underreporting. In a randomized, stratified community survey conducted in the U.S. Virgin Islands, the calculated incidence rate was 73 cases/10,000 population/year (Morris et al., 1982b). In Puerto Rico, 45 cases were reported to the Puerto Rico Poison Control Center in 1982. In an associated telephone survey, 7% of persons contacted reported that at least one family member had at one time had ciguatera (Holt et al., 1984). In Miami, 129 cases of ciguatera were reported to the Dade County Health Department between 1972 and 1976, for an annual incidence of 5 cases/100,000 population; the actual incidence was estimated to be 10-100 times this figure (i.e., 50-500 cases/100,000 population) (Lawrence et al., 1980). An incidence rate of 234.9 cases/100,000 population was reported for the Marshall Islands during 1982-1987 (Ruff, 1989). The average annual incidence in Hawaii for 1984-1988 was only 8.7 cases/100,000 population, but this varied greatly from island to island: in 1988 the rate per 100,000 population was 3.2 on Oahu, 12.5 on Kauai, 11.1 on Maui, 33.9 on Hawaii (the largest island), and 7.5 for the state (Gollop and Pon, 1991). These data emphasize the striking regional nature of this disease and its very real importance as a cause of morbidity in endemic areas. There is some evidence from the Pacific region that changes in the reef environment due to construction or other underwater activities can cause an increase in the occurrence of ciguatoxic fish (Anderson et al., 1983; Ruff, 1989).

TABLE 4-3

Outbreaks and Cases of Ciguatera in the United States Reported to CDC 1978-1987.

For the vast majority of U.S. consumers, the disease can be contracted only through consumption of fish imported from endemic areas. For residents of south Florida, the Caribbean, and Hawaiian or other Pacific islands, absolute safety depends on individual abstinence from eating reef fish. The risk may be greatly reduced by avoidance of particular species of fish from known "hot-spot" areas (Randall, 1980). However, hot spots may persist for extended periods of time or may change (Cooper, 1964; Halstead, 1967). Voluntary action by commercial fish distributors in some areas has been quite effective. Thus amberjack (Seriola dumerili, Kahala) is not sold commercially in Hawaii because of the known high incidence of ciguatoxic fish of this species. Other suspect species coming to market in Hawaii may be tested by the "stick test." Also, the Hawaii Department of Health (HDH, 1988) publishes a warning pamphlet on Fish Poisoning in Hawaii and periodically issues advisories on dangerous species and the areas from which they have been taken.

In areas where reef fish are part of the regular diet of inhabitants or visitors, and particularly where significant quantities of fish are caught and consumed by recreational or small-boat fishermen, it seems unlikely that ciguatera can be completely prevented. The impossibility of detecting toxic fish by organoleptic inspection and the sporadic occurrence of such fish limit control options. The availability of a simple reliable test would greatly improve the situation. At present, the only ciguatera screening program in existence is that employed by the Tokyo Central Wholesale Fish Market in Japan. Hygiene inspectors examine incoming shiploads of fish that have originated in tropical island regions. Suspect specimens are removed for testing. Muscle extracts are prepared and tested on cats and mice for evidence of ciguatoxicity (Halstead, 1970). This is a lengthy and expensive screening technique that is impractical when dealing with large numbers of samples. A radioimmunoassay (RIA) was developed by Hokama and co-workers in Hawaii (Hokama et al., 1977) and then modified to a simpler, enzyme immunoassay (Hokama, 1985). The method has been further simplified to a "stick" test that has been used to screen fish landed in Hawaii and holds promise as a practical basis for control (Hokama et al., 1989b).

However, even this test costs $1 to 2 per fish, and it would not be possible to test each reef fish landed. Kits are being developed for use by sports fishers that could partly resolve the cost problem. In any case it would be desirable to limit testing to high-risk fish. Research is needed into methods for predicting the development of ciguateric conditions in reef fishing areas, perhaps by assessing Gambierdiscus or other toxigenic microorganism populations and somehow closing such areas to fishing when the risk is high. Reef closure would probably be feasible in discrete Pacific islands where there is limited movement of fish from one reef area to another. However, this might not be the case in the Caribbean where fish movement between reefs is easier and more common. Obviously, there should be some follow-up on the reports of a relationship between reef disturbance and increased occurrence of ciguateric fish because this may result from human activity that can be controlled (Gollop and Pon, 1991).

Estimates of the economic consequences of ciguatera are not easily made. However, they are significant for island communities largely dependent on tourism. Ragelis (1984) quoted an estimate of an annual loss to fishermen in the Caribbean region and Hawaii of $10 million as a result of restricted fishing, but this may be low.

In summary, the risks of contracting ciguatera fish poisoning are low for most consumers of seafood in the mainland United States. Risks are much higher in Hawaii, other Pacific islands, Puerto Rico, and the Virgin Islands, with more moderate risks in areas such as Miami that border endemic zones. For mainland consumers, protection could be afforded by strict control of imports and intrastate shipments. However, such an approach may be unnecessarily severe and does not address the much more significant problems that exist in endemic areas. A more reasonable (and potentially cost-effective) approach would be to emphasize development of an inexpensive but reliable assay for ciguatoxic fish, similar to the stick test proposed by Hokama (1990). The stick test measures ciguatoxin and polyether compounds including okadaic acid (Hokama et al., 1989a). In a recent outbreak in Hawaii due to Philippine fish, the test was positive for a fish that was then shown to contain palytoxin (Kodama et al., 1989). It has been suggested that palytoxin, previously reported from parrot fish and crab in Japan, is one of the toxins "under the rubric of ciguatera" (Kodama et al., 1989). Obviously, this whole area needs further research, particularly because of the concern over false-positive results from the stick test. If such an assay were widely available, it might be applied both by regulation and voluntarily to reduce the incidence of disease in endemic areas, particularly if consumers were knowledgeable and insisted on purchasing only fish that had been screened for toxicity. Similarly, interstate shipment and imports of potentially high-risk fish (grouper, jack) could be restricted to fish certified to be nontoxic. This is particularly important in view of the increased production and export of fish to the United States from Pacific islands and reef fishery areas of Southeast Asia, such as the Philippines (Miller, 1991).

Scombroid (Histamine) Fish Poisoning

Scombroid intoxication results from ingestion of fish containing high levels of free histamine. Initially, the disease was associated with consumption of scombroid fish such as tuna, mackerel, bonito, and saury. More recently, other types of fish have been identified as causing the intoxication, including mahimahi, bluefish, jack, mackerel, amberjack, herring, sardine, and anchovy. In the United States, scombroid fish poisoning has been caused dominantly by mahimahi, tuna, and bluefish (CDC, 1989) (see Table 4-4).

TABLE 4-4

Fish Reported to CDC as Vehicles of Scombroid Poisoning in the United States, 1978-1986.

Scombroid food poisoning has a wider geographic occurrence in the United States than ciguatera, with incidents reported from 45 states during 1978-1988. The highest number of outbreaks (45) and cases (171) occurred in Hawaii, but mainland states reported a total of 111 outbreaks and 582 cases (see Table 4-5). This reflects the fact that the disease, although associated with warm ambient temperatures, is not due solely to tropical or subtropical species of fish. Thus, the risk of scombroid poisoning is widespread among fish-eating consumers. Fortunately, the disease is mild, of short duration, and self-resolving without any sequelae in the vast majority of cases. Moreover, because the toxic condition is a consequence of improper handling or storage of the fish and there are effective testing methods to identify toxic fish, control and prevention are possible. The mildness and transient nature of scombroid poisoning make it likely that this disease is underreported.

TABLE 4-5

Scombroid Fish Poisoning in the United States Reported to CDC, 1978-1987.

Fish imported to the United States from warmwater countries, particularly mahimahi, have been implicated as a cause of scombroid poisoning; this reflects both the high ambient water and air temperatures in the originating area, and the poor handling conditions on boats and in markets permitting growth of the bacteria that convert histidine to histamine.

The disease is correctly described as histamine poisoning (Taylor, 1986); it includes gastrointestinal, neurological, hemodynamic, and cutaneous symptoms such as nausea, vomiting, diarrhea, cramping, headache, palpitations, flushing, tingling, burning, itching, hypotension, rash, urticaria, edema, and localized inflammation. The most frequent symptoms are tingling and burning sensations around the mouth ("peppery tasting"), gastrointestinal complaints, and a rash with itching. The illness is generally mild and self-resolving, with rapid onset of symptoms and duration of only a few hours. Normally, treatment is unnecessary but antihistamine drugs will provide relief.

The histamine is produced in the fish flesh by decarboxylation of free histidine, which is naturally present at high levels in species of fish implicated in scombroid fish poisoning (Lukton and Olcott, 1958). The production of histamine is due to the action of histidine decarboxylase, an enzyme produced by bacteria growing on the fish. Histidine decarboxylase production is not widespread among bacteria and is found principally among species of the Enterobacteriaceae, Clostridium, Lactobacillus (Taylor, 1986), and possibly Vibrio (Van Spreekens, 1987). The enteric bacteria Morganella morganii, Klebsiella pneumoniae, and Hafnia alvei have been isolated and identified from fish implicated in histamine poisoning (Havelka, 1967; Kawabata et al., 1956; Taylor et al., 1979). Other enteric bacteria, Clostridium perfringens, and halophilic vibrios have also been reported, but M. morganii and K. pneumoniae are most frequently implicated. These organisms are not commonly isolated from living fish and may be added during catching and handling (Taylor, 1986).

Bacteria must grow to a large enough population for significant production of histamine to occur. These are mesophilic bacteria that require temperatures higher than 15°C. In tropical areas of the world, fish temperatures at capture frequently exceed 20°C, and on small vessels it is not unusual for fish to be held on deck at even higher temperatures for several hours. Histamine production is optimal at 30°C (Arnold et al., 1980). Once a large population of bacteria has been established, residual enzyme activity continues slowly at refrigeration temperatures (0-5°C) though bacterial growth ceases.

Thus, histamine production in fish is a consequence of improper handling and storage of fish after capture. Indeed, histamine content may be used as an index of spoilage in certain fish. The Food and Drug Administration (FDA) considers a level of 20 milligrams (mg) of histamine per 100 grams (g) of flesh, or 200 parts per million (ppm), an indication of spoilage in tuna and 50 mg/100 g (500 ppm) an indication of hazard (Federal Register, 1982). This is close to the toxic dose estimate of 60 mg/100 g made by Simidu and Hibiku (1955). There is uncertainty regarding the threshold toxic dose because potentiators of toxicity are present in fish that lower the effective dosage compared with pure histamine.

The occurrence of scombroid fish poisoning in recent years, based on CDC reports, is between 12 and 20 outbreaks involving fewer than 100 cases per year (higher numbers were recorded in 1973, 1979, and 1980). This is, without question, a considerable underestimate because the illness is generally mild, passes rapidly with no after effects, and is thus not usually reported to health authorities. Good chemical tests are available for histamine in fish flesh (Taylor, 1986), which has allowed FDA to set an action level for histamine in tuna at 50 mg/100 g of flesh. Above this level the fish is considered hazardous. Fish histamine poisoning is preventable by proper handling of fish at the time of capture and during subsequent storage, processing, and distribution. Fish should be chilled as rapidly as possible after capture by using ice, refrigerated seawater or brine, or mechanical refrigeration. Flesh temperature should be brought below 15°C and preferably below 10°C within 4 hours; this should be normal practice in commercial systems. Histamine levels should be monitored routinely by the industry in susceptible species where proper prior handling cannot be ensured. The level at which testing is performed will depend on the species and product form (e.g., tuna for canning, hot smoked mackerel). In the United States, the highest-risk fish commercially is probably imported fresh or frozen fish from tropical areas. High histamine levels may be present in such fish when other overt signs of spoilage (bad odor, discoloration) are absent. Imported fish should be subject to controls. Domestically caught species in normal commercial channels are probably less of a problem because of the widespread use of ice and refrigeration. Bluefish and sport caught tuna or mackerel present a more intractable control problem because they are caught by individuals and either do not enter commercial channels or do so in an unconventional way. Some local or state control may be possible where licensed charter fishing boats are involved, perhaps by requiring that adequate facilities are provided for rapid chilling of fish and for their storage in a chilled state until landed. In the absence of a simple litmus test, control for most sports fishers and their families depends on education. States should be encouraged to provide advisory bulletins to sports fishers. However, it should be emphasized that this is a mild disease that is neither long lasting nor life threatening, and that symptoms can be relieved quickly by antihistamines.

Paralytic Shellfish Poisoning (PSP)

Paralytic shellfish poisoning results from ingesting bivalve molluscs (mussels, clams, oysters, scallops) that have consumed toxigenic dinoflagellates (Halstead and Schantz, 1984; Schantz, 1973). The toxins are assimilated and temporarily stored by the shellfish. In the United States, PSP is a problem primarily in the New England states on the East Coast and in Alaska, California, and Washington on the West Coast. Very few outbreaks have occurred in other areas of the United States from shellfish harvested in coastal states, reflecting the effectiveness of current testing and control measures for commercially produced shellfish. Most disease incidents involve mussels, clams, and scallops gathered and eaten by recreational collectors often from closed areas. The CDC listed 12 outbreaks involving 134 people with one death during 1978-1986 (Table 4-6). The outbreaks occurred in Alaska, California, Massachusetts, Tennessee (due to mussels from California), and Washington (Table 4-7). The Northeast Technical Support Unit (NETSU) compilation shows a total of 282 cases for the period, including cases from Maine, Alaska, and Massachusetts. A report by Nishitani and Chew (1988), based on data from the West Coast, lists cases as follows: 68 from Alaska, 98 from California, 1 from Oregon, and 12 from Washington. Thus, there is some evidence of underreporting of cases to CDC. Although PSP is an extremely dangerous disease that can cause death, there is reason to believe that mild cases due to consumption of marginally toxic clams by recreational diggers are never reported to health authorities or are misdiagnosed.

TABLE 4-6

PSP Incidents in the United States Reported by CDC, 1978-1985.

TABLE 4-7

PSP Incidents in the United States Listed Elsewhere.

Paralytic shellfish poisoning is potentially life threatening because the toxins involved are among the most poisonous known. Symptoms are neurological and normally appear within an hour of eating toxic shellfish; in nonlethal cases they usually subside within a few days. Symptoms include tingling, numbness, and burning of the lips and fingertips; ataxia; giddiness; staggering; drowsiness; dry throat and skin; incoherence; aphasia; rash; and fever. In severe cases, respiratory paralysis occurs, which can cause death usually during the first 24 hours, so that the prognosis for recovery is good for patients surviving this period. No antidote is known, but respiratory support is given when paralysis occurs. There are no sequelae, and patients recover completely. Immunity is not conferred by a poisonous episode and multiple incidents can occur.

The cause of PSP is a complex of toxins known as saxitoxins because all can be considered forms or derivatives of saxitoxin, whose structure was reported by Schantz et al. (1975). The 12 most commonly encountered include saxitoxin, neosaxitoxin, gonyautoxins (I, II, III, IV), B1, B2, C1, C2, C3, and C4, which vary in their toxic effects on mice (Boyer et al., 1978; Shimizu and Hsu, 1981). Saxitoxin, neosaxitoxin, and gonyautoxins II and III are roughly equal in toxicity, whereas the others are somewhat weaker (Hall and Reichardt, 1984). In the United States, the toxigenic dinoflagellates of importance are Gonyaulax catenella and G. tamarenses,¹ the first being most dominant on the West Coast and the second on the East Coast (Taylor, 1988). Strains of these microorganisms develop characteristic toxin profiles that usually contain six to eight saxitoxins. Shellfish feeding on blooms of these Gonyaulax ingest all toxins but seem to selectively retain or biologically modify some derivatives because the toxin profiles in clams or scallops may differ from those of the Gonyaulax on which they have been feeding (Schantz et al., 1975; Sullivan et al., 1983).

The saxitoxins are neurotoxins that act by blocking the flow of sodium (Na⁺) ions through the sodium channels of nerves, thereby interfering with signal transmission. The lethal dose for humans is 1-4 mg expressed as saxitoxin equivalents (Schantz, 1986), and the FDA action limit is 80 micrograms (µg) of toxin per 100 g of shellfish tissue.

The classical method for analysis of saxitoxins is by mouse bioassay (Schantz et al., 1958), a standardized procedure in which the strain, size, and condition of the mouse are all important. Suitable groups of mice are challenged by injection of toxin extracts, and their responses are compared with those of mice injected with known toxin at different concentrations. This method is used by state laboratories in surveillance activities. The chemical analytical methods that have been used, mostly in research, include column separation, thin-layer chromatography, and fluorescent assays either directly or after separation by high-performance liquid chromatography (HPLC). The HPLC method is quite rapid and has been considered as a possible replacement of the mouse bioassay (Sullivan and Iwaoka, 1983). Recently, immunoassay methods have been developed and reported. These involve rabbit serum antibody preparations and monoclonal antibodies, and both radioimmunoassays and enzyme-linked immunoabsorbent assays (ELISA) have been used (Chu and Fan, 1985). However, so far these methods are not acceptable for regulatory use.

Blooms of the toxigenic dinoflagellates Gonyaulax catenella (West Coast) and G. tamarensis (East Coast) occur several times each year, primarily from April to October along the U.S. West Coast, Alaska, and the East Coast from Long Island Sound through Maine. Similar blooms occur off British Columbia and the Canadian Maritime Provinces. The occurrence of blooms is not predictable with the present state of knowledge. When these blooms occur, shellfish become toxic and remain toxic for several weeks after the bloom subsides. Some species of bivalve molluscs in a few areas remain constantly toxic (e.g., butter clams in parts of Washington State and in Alaska). Protection of the consumer is primarily the responsibility of the state in which potentially toxic shellfish originate and is achieved by closure of shellfish harvesting in affected areas. Closure may be absolute for certain species over a long period (e.g., ocean coast mussels in California) or temporary (e.g., most hard-shell clam operations in Washington and Oregon) (Nishitani and Chew, 1988). Because clam digging and oyster gathering from public beaches are popular recreational activities in coastal states, the authorities issue warnings through the media when dangerous conditions exist and post warning notices (multilingual in the West) on public beaches. Each of the affected states has a surveillance system involving regular sampling and testing of shellfish from different areas during the season for blooms (April to October). Toxicity tests are normally run in state public health laboratories by using the standardized mouse bioassay (AOAC, 1984), whereas the collection of samples and posting of beaches are frequently the responsibility of local (e.g., county) authorities. Commercial producers are required to submit samples for testing, and many voluntarily send shellfish to the state laboratories.

This system, although expensive for the states, has served the U.S. consumer well, which can be seen from the very small number of outbreaks reported to CDC in recent years. In 1977-1986, these involved only twelve outbreaks with 134 cases. Only two outbreaks, both due to mussels from the same source (California), were of commercial origin. All other outbreaks were among recreational clam diggers (shellfish gatherers). Despite the seriousness of the intoxication, there have been only two deaths in the last 10 years. West Coast state records show a somewhat higher number of outbreaks, which have typically involved recreational shellfish gatherers but with no serious consequences (Nishitani and Chew, 1988).

The present system of testing and control by the states appears to provide adequate protection to the consumer of domestically produced shellfish.

Neurotoxic Shellfish Poisoning (NSP)

Neurotoxic shellfish poisoning (sometimes referred to as brevetoxic shellfish poisoning, BSP) is caused by ingesting shellfish that have fed on the red tide organism Gymnodinium breve (formerly Ptychodiscus brevis). Red tides occur sporadically in the Gulf of Mexico and off the coast of Florida (Baden et al., 1984). They may be carried north in the Gulf Stream, occasionally affecting the coastline of adjacent states. The dinoflagellate blooms are easily observed as a red coloration of seawater, and the organisms can be detected microscopically. Red tides usually cause massive fish kills, and the carcasses wash ashore. Irritant aerosols are produced by wind and wave action, which may cause respiratory distress. Filter feeding molluscs ingest the dinoflagellates and retain the toxin in their tissues for some time. No cases of NSP were reported to CDC in the period 1978-1986, but five cases were reported from Florida in 1973-1974 according to NETSU. During a red tide incident in North Carolina in 1987-1988, 48 persons became ill with NSP (Tester and Fowler, 1990). Reports of respiratory irritation affecting people in coastal areas were received on October 29 and 30, 1987; oyster harvesting was closed on November 2. However, 35 of the reported cases occurred prior to the ban on shellfish harvesting which lasted from 3 1/2-6 months, depending on location. The basis for closure was the occurrence of more than 5,000 G. breve cells per liter of seawater, and reopening of harvest was dependent on demonstrated absence of brevetoxin in 100-g samples of shellfish meat based on a mouse bioassay (DNR, 1985; FDA, 1989). Notwithstanding this incident, which occurred as a result of the first bloom of G. breve reported in North Carolina, the general record in recent years suggests that the surveillance and closure systems operated by the states are indeed effective.

Symptoms resulting from the ingestion of shellfish containing brevetoxins include tingling and numbness of the lips, tongue, throat, and perioral area; muscular aches; gastrointestinal upset; and dizziness. The intoxication is usually not fatal. Onset is rapid and symptoms subside within a few hours or days at most. There is no antidote.

The symptoms appear to be due to two brevetoxins produced by G. breve that bind to nerve cells (Baden et al., 1984). They may be analyzed chemically, but this is not done routinely. Identification of a dangerous condition is readily made by observation of red tide conditions, including characteristic fish kills, and of the organisms themselves in the water. Local authorities then routinely close shellfish harvesting to industries and the public.

Diarrhetic Shellfish Poisoning (DSP)

Diarrhetic shellfish poisoning is caused by ingestion of mussels, scallops, or clams that have been feeding on Dinophysis fortii or D. acuminata and other species of Dinophysis and possibly Prorocentrum (Edler and Hageltorn, 1990; Yasumoto and Murata, 1990). There have been no confirmed outbreaks in the United States, but the disease is common in Japan and has become a problem in Europe. One confirmed DSP episode occurred in Canada in 1990.

Symptoms include diarrhea, nausea, vomiting, and abdominal pain. Onset occurs from 30 minutes to a few hours after eating toxic shellfish, and the duration is usually short with a maximum of a few days in severe cases. The disease is not life threatening (Yasumoto et al., 1984).

At least five toxins have been isolated from dinoflagellates and shellfish. Okadaic acid is most commonly encountered in Europe where D. acuminata is the usual agent, and mixtures of okadaic acid, dinophysistoxins, and pectenotoxins are detected in Japanese cases usually involving D. fortii (Yasumoto and Murata, 1990). There is a mouse bioassay for the toxins.

For the U.S. consumer, DSP would appear, at present, to be a hazard only for imported products and should be controllable by import regulation. Shellfish should be imported only from countries with whom the United States has a memorandum of understanding (MOU). Testing for shellfish toxins should be part of the general practice under the MOU. Nevertheless, because Dinophysis does occur in U.S. coastal waters, regulatory agencies in the United States should be alert for the possibility of an outbreak (Freudenthal and Jijina, 1988).

Puffer Fish Poisoning (PFP)

Puffer fish poisoning results from ingestion of the flesh of certain species of fish belonging to the Tetraodontidae (Halstead, 1967). The toxin involved is called tetrodotoxin and was originally believed to be a true ichthyosarcotoxin produced by the fish itself. The toxicity of poisonous puffers fluctuates greatly (Halstead, 1988). Recent observations that cultured puffer fish are atoxic has supported a food chain origin for the toxin, but this has not yet been confirmed (Mosher and Fuhrman, 1984). It has recently been shown, however, that certain common marine vibrios can produce a form of the toxin (Narita et al., 1987), and because vibrios occur as part of the microflora of puffer fish, they may be implicated in toxicity development (Sugita et al., 1989).

Puffer fish poisoning has not been reported in mainland United States in recent years, but incidents were reported in the past. Seven cases were reported in Florida between 1951 and 1974, including three fatalities (Benson, 1956; Hemmert, 1974). They appear to have been caused by the consumption of locally caught species of Sphoeroides. The common puffer fish, Arothron hispidus, has been implicated in at least seven fatalities in Hawaii (HDH, 1988). There are 20-100 deaths from fugu poisoning in Japan each year, where various species of puffer fish are eaten as a delicacy; this occurs despite very stringent controls imposed by Japanese authorities on the marketing and restaurant preparation of the dish (Ogura, 1971).

The symptoms of puffer fish poisoning are similar to those described for paralytic shellfish poisoning, including initial tingling and numbness of lips, tongue, and fingers leading to paralysis of the extremities; ataxia; difficulty in speaking; and finally, death by asphyxiation due to respiratory paralysis. Nausea and vomiting are common early symptoms. The similarity in symptoms is not surprising because tetrodotoxin, although chemically different from the saxitoxins, also blocks sodium channels. No antidote has been identified for tetrodotoxin and treatment is supportive. The toxicity of tetrodotoxin is similar to that of saxitoxin, and 1-4 mg constitutes a lethal dose for humans.

There is disagreement concerning the toxicity of U.S. Atlantic puffer fish. A recent advisory from the National Oceanic and Atmospheric Administration (NOAA, 1988) describes the northern puffer (Sphoeroides maculatus) as nontoxic and notes that the fish were marketed along the Atlantic Coast as "sea squab" during World War II. However, Hemmert (1974) shows a table indicating that the viscera, skin, and some flesh of S. maculatus caught in the Atlantic were toxic (Lalone et al., 1963). Larson et al. (1959, 1960) also reported that S. maculatus is toxic. From the West Coast, Goe and Halstead (1953) reported that the Pacific species S. annulatus is often toxic. The species Arothron hispidus has been implicated in at least seven fatalities in Hawaii. The wholesaling, preparation, and selling of puffers as food in Japan, even under the most rigid public health conditions by trained and certified puffer cooks, has not eliminated the danger of eating these fish. The fugu (puffer) still remains a major cause of fatal food intoxications in Japan. In brief, eating poisonous puffers is at best a game of Russian roulette. All of the U.S. puffers may be potentially toxic. There are too many variables in the puffer business, and sale of these should be prohibited in the United States. This subject has been documented and discussed at great length by Halstead (1967, 1988).

In view of these reports, it would seem prudent to exclude puffer fish, whether domestic or imported, from U.S. commercial channels at least until a proper assessment is made of the extent of risk they may present. The FDA has recently approved the importation of the Japanese puffer for fugu restaurants in the United States. Even though very strict requirements have been imposed in an attempt to ensure that the fish are nontoxic, the continuing Japanese experience should raise questions concerning the safety of this process for the U.S. public (Halstead, 1988).

Amnesic Shellfish Poisoning (ASP)

Amnesic shellfish poisoning has been proposed by Todd (1989) as a name for the syndrome caused by domoic acid. This severe disease has been identified only in a series of outbreaks in Canada in November and December 1988 involving 103 people. The toxin is present in some varieties of the diatom Nitzschia pungens and accumulated in mussels and clams in Atlantic Canada during a period of blooms of the diatom. Symptoms included vomiting, abdominal cramps, diarrhea, disorientation, and memory loss (Perl et al., 1988; Teitelbaum et al., 1990). Short-term memory loss was the most persistent symptom and lasted over a year in several cases. Autopsies on three fatalities showed necrosis of the hippocampus. The disease is particularly severe among older people, some of whom died in the Canadian outbreak.

Canadian authorities now analyze mussels and clams for domoic acid and enforce closure of beds when levels in excess of 20 µg/g are detected in their tissues (Gilgan et al., 1989).

Clearly, this is a toxin to be considered in U.S. testing regimes, and there should be close cooperation between U.S. and Canadian regulatory agencies on the movement of imported Canadian shellfish into the United States. Nitzschia pungens and N. pseudodelicatissima reportedly occur in northern U.S. and Canadian waters, and there is potential for development of toxicity in shellfish growing in these areas. States in the northeastern United States are now testing mussels for domoic acid.

Other Toxins

There are sporadic reports of other intoxications from seafoods from time to time (Halstead, 1988; Wekell and Liston, 1982), but these have not been investigated sufficiently to identify the toxic agent. As noted earlier, the somewhat variable symptoms defined as ciguatera and the reported association of polyether substances and palytoxin (Hokama et al., 1989a; Kodama et al., 1989) in fish implicated in such cases raise questions about the toxicity of reef-associated fish.

One well-defined syndrome reported to occur in Hawaii is "hallucinogenic fish poisoning." This illness follows consumption of mullet and a number of reef fish and occurs seasonally, usually during summer months. Hallucinations, insomnia, intense dreaming, weakness, and burning of the throat are common soon after eating the fish (Halstead and Schantz, 1984). Terrifying nightmares have been reported and constrictive chest pains occur. The condition is short-lived and self-resolving (Halstead, 1988; HDH, 1988). There does not seem to be any analytical test for this toxin.