|Home | About | Journals | Submit | Contact Us | Français|
Gnathostomiasis is a food-borne zoonosis caused by the late-third stage larvae of Gnathostoma spp. It is being seen with increasing frequency in countries where it is not endemic and should be regarded as another emerging imported disease. Previously, its foci of endemicity have been confined to Southeast Asia and Central and South America, but its geographical boundaries appear to be increasing, with recent reports of infection in tourists returning from southern Africa. It has a complex life cycle involving at least two intermediate hosts, with humans being accidental hosts in which the larvae cannot reach sexual maturity. The main risks for acquisition are consumption of raw or undercooked freshwater fish and geographical exposure. Infection results in initial nonspecific symptoms followed by cutaneous and/or visceral larva migrans, with the latter carrying high morbidity and mortality rates if there is central nervous system involvement. We review the literature and describe the epidemiology, life cycle, clinical features, diagnosis, treatment, and prevention of gnathostomiasis.
International travel to the tropics has dramatically increased over the past few decades, with a subsequent and significant increase in the number of patients presenting with tropical diseases in countries where such infections are not endemic. It is estimated that 50 million residents of industrialized countries travel annually to such areas (42, 44), which brings exposure to a broad range of pathogens rarely, if ever, encountered at home. These may vary from short-lived, easily detected and treatable infections (e.g., gastrointestinal infections) to more exotic infections such as filarial and helminthic infections (e.g., loiasis, stronglyoidiasis, and schistosomiasis). Such infections may be seen rarely by physicians in temperate climates, and therefore diagnosis can prove elusive if these infections are not considered. Travelers are becoming ever more adventurous in choice of country, pursuit of remoteness and immersion in local culture, which will frequently include eating all local delicacies (e.g., ceviche or “drunken crab”) without consideration of what organisms they might be harboring. Migration has also increased substantially over the past few decades, with people from the tropics and subtropics settling in the West, and many come harboring parasites of which they are unaware.
Gnathostomiasis is a parasitic infection caused by the third-stage larvae of the helminths Gnathostoma spp., which are seen mostly in tropical and subtropical regions. It is a food-borne zoonosis and is endemic in areas where people eat raw freshwater fish or shellfish, especially Thailand and other parts of Southeast Asia, Japan, and increasingly Latin America, particularly Mexico. Previously, the disease was rarely seen outside areas of endemicity; however, over the past decade, the number of cases seen in countries where it is not endemic has increased. Few clinicians outside areas of endemicity are familiar with the disease, and therefore diagnosis is often missed or prolonged, with potentially serious consequences. The classic triad of intermittent migratory swellings, eosinophilia, and a history of travel to Southeast Asia or other areas of endemicity should alert physicians to the possible diagnosis. Visceral disease is more serious than the cutaneous manifestations and, in the case of central nervous system (CNS) disease, may be fatal.
This article describes the epidemiology, life cycle, clinical features, diagnostic tools, treatment, and prevention of this disease.
Gnathostoma was first discovered in the stomach wall of a tiger that died at London Zoo in 1836 (35) and was first described in humans in 1889 in Thailand by G. M. R. Levinson (cited in references 5 and 24). The next case was not described until 1934, and shortly after, its life cycle was elucidated (37).
The foci of endemicity have been predominantly in Japan and Southeast Asia, particularly Thailand, but the disease is also endemic in Cambodia, Laos, Myanmar, Indonesia, Philippines, and Malaysia (Fig. (Fig.1).1). Cases have also been reported in China, Sri Lanka, and India (41). In more recent years it has become an increasing problem in Central and South America, particularly in Mexico (due to the consumption of ceviche [raw fish marinated in lime]) (12, 39), and also in Guatemala, Peru, and Ecuador (14, 23). There have also reports of cases in Myanmar, Zambia, and, most recently, Botswana (6, 16, 17). Changes in dietary habits are the main cause of expansion of the geographical range of the disease. However, despite the significant increase in the number of sushi bars in the West, there do not appear to be any data showing a causal increase in the disease. From the reports, it seems that the important factor is where the sushi is eaten rather than simply the consumption itself. Nawa et al. (28) have suggested that the cases tend to occur as a result of consumption of food from local restaurants in countries where the disease is endemic and where few regulations if any govern the sourcing or storage of fish for consumption. They suggest that such restaurants tend to use cheaper local freshwater or brackish-water fish, in contrast to sushi bars and restaurants in the West, which primarily use more expensive saltwater fish which are free of Gnathostoma spp. and harbor relatively few potentially transmissible parasites. In the latter places there are very stringent rules governing food sourcing and preparation which have to be adhered to and are rigorously enforced (28).
The genus Gnathostoma belongs to the order Spirurida, one of the largest groups of nematodes. These groups are characterized biologically by requiring one or more intermediate hosts in their life cycles. The genus has 12 species (11, 15, 29), with only 4 recorded in humans: G. spinigerum, commonly found in wild and domestic cats and dogs in India, China, Japan, and southeast Asia; G. hispidum, found in wild and domestic pigs in Europe, Asia, and Australia; G. doloresi, found in wild boars; and G. nipponicum, found in weasels in Japan (15). Until the early 1980s, all cases of gnathostomiasis found in humans in the areas of endemicity in Thailand, China, and Japan were solely due to G. spinigerum. However, in the early 1980s in Japan new cases appearing in urban areas were found to be due to G. hispidum, with the infective larvae found in loaches (small freshwater fish) (29).
Humans are accidental hosts in which the parasite fails to reach sexual maturity. The definitive hosts are dogs, cats, tigers, leopards, and probably other fish-eating mammals, where the adult worm lives coiled up in the wall of the stomach, producing a tumor-like mass. In the definitive hosts the adult worm (which reaches 13 to 55 mm in length) releases eggs into the stomach which are then passed in the feces. These embryonate in freshwater and hatch to release first-stage larvae after about 7 days, which are then ingested by the first intermediate host, a water flea or copepod (usually a species of the genus Cyclops), where they develop into second-stage larvae (Fig. (Fig.2).2). When infected Cyclops organisms are ingested by the second intermediate hosts (fish, eels, frogs, birds, and reptiles), the second-stage larvae are freed in the intestine and develop into third-stage larvae. These migrate through the tissues and encyst in the muscles of their transport hosts, where they remain as infectious larvae. When eaten by an appropriate definitive host, such as cats and dogs, the larvae are freed once more in the gastrointestinal tract, from where they migrate to the liver and abdominal cavity. After about 4 weeks they return to the stomach, invading the gastric wall, where they resemble a tumor, with an aperture that connects to the gastric lumen through which eggs may later be released. Here they grow into adults to complete their life cycle, usually within about 6 months. Eggs are released into the environment in the stool of the host about 8 to 12 months after initial ingestion of the infective third-stage larvae by the definitive host (15, 41, 51). Humans usually become infected with the third-stage larvae of Gnathostoma spp. by eating raw or inadequately cooked freshwater fish or other intermediate hosts such as snakes, frogs, and chickens. However, two alternative routes of infection have been suggested: ingestion of water containing infected copepods (thus taking the place of a second intermediate host) or by penetration of the skin of food handlers by third-stage larvae from infected meat (11). Symptoms in humans occur as the late third-stage larvae migrate through the tissues, causing intermittent symptoms of cutaneous or visceral larva migrans. The larvae have been observed to move at 1 cm/hour (14).
The larvae usually measure up to 12.5 mm long and 1.2 mm wide and are reddish white in color (15). They are identified on the basis of several characteristic features, which include the shape of the body, the number of rows of hooks at the cephalic end, the number of hooks in each row, the character of the spines which cover the body, and the extent to which the body is covered by spines, which vary according to species (Fig. (Fig.3)3) (11). These rows of hooks enable the larva to lodge in the host tissues and are in part responsible for the mechanical damage that it inflicts on its host. Identification of the species present is important for epidemiological purposes, but treatment is the same for all species.
The clinical features can be divided into immediate symptoms, a cutaneous form, and a visceral form. Within 24 to 48 h of ingestion of Gnathostoma organisms, patients may develop nonspecific signs and symptoms such as malaise, fever, urticaria, anorexia, nausea, vomiting, diarrhea, and epigastric or right upper quadrant pain. These symptoms occur as the larva excysts and migrates through the stomach or intestinal wall and the liver and may last for 2 to 3 weeks (15). A marked generalized eosinophilia usually develops in association with larval penetration of the gastrointestinal wall, with reported levels of >50% of the total white cell count. The larval worm then migrates to the skin through the subcutaneous tissue causing the typical migratory swellings (cutaneous disease) and from here may penetrate into deeper tissues and viscera to involve the lungs, eyes, ears, gastrointestinal and genitourinary systems, and rarely, but often fatally, the CNS (visceral disease). The majority of infections result only in cutaneous disease, within 3 to 4 weeks after ingestion of the larvae, but the onset of symptoms may be delayed for months and even years (23, 41). As the chronic stage begins and the larva enters the subcutaneous tissues, the eosinophilia and systemic features usually subside.
The exact pathogenicity of gnathostomiasis is uncertain, but it is thought that the symptoms are due to the combined effects of mechanical damage secondary to the larva's migration, the excretions and secretions it produces, and the host's immunological response. The substances released contain various compounds, including one similar to acetylcholine, a “spreading factor” with hyaluronidase, a proteolytic enzyme, and a hemolytic substance, which have been demonstrated in various studies in Japan in the 1950s (5, 24). These substances, in addition to the mechanical damage, result in the characteristic hemorrhagic tracks that may be seen in the subcutaneous tissues in patients or in the viscera or CNS postmortem.
Cutaneous gnathostomiasis is the most common manifestation of infection and is known by several local names, e.g., Yangtze River's edema and Shanghai's rheumatism in China, tuao chid in Japan, and paniculitis nodular migratoria eosinofilica in Latin America. It typically presents with intermittent migratory swellings, (nodular migratory panniculitis), usually affecting the trunk or upper limbs. These nonpitting edematous swellings vary in size and may be pruritic, painful, or erythematous (Fig. (Fig.44 and and5).5). They usually occur within 3 to 4 weeks of ingestion of the larvae, typically last 1 to 2 weeks, and are commonly due to only one larva, but on occasion infection with two or more has been found (10, 41). The swellings are due to both mechanical damage from the larva and the host's immunological response to the parasite and its secretions. As the larva migrates, subcutaneous hemorrhages may be seen along its tracks, which are pathognomonic of gnathostomiasis and can help differentiate it from other causes of larva migrans, e.g., sparganosis or strongyloidiasis. Episodes of swelling slowly become less intense and shorter in duration, but in untreated patients symptoms may recur intermittently for up to 10 to 12 years.
Other, less common manifestations of cutaneous gnathostomiasis include a creeping eruption (which may be confused with cutaneous larva migrans), a skin abscess, or a nodule, which tend to occur when the larva is migrating more superficially (41). In these cases the larva can often be excised. Spontaneous extrusion of a larva from the subcutaneous tissues has also been described. If the migratory lesions are on the face, there is a serious risk of CNS or ocular invasion.
The Gnathostoma larva is highly invasive and motile and therefore can produce an extremely wide range of symptoms affecting virtually any part of the body. In noncerebral disease the larvae may continue to cause intermittent symptoms until they die after about 12 years, if left untreated.
Pulmonary symptoms that have been attributed to infection with Gnathostoma spp. include cough, pleuritic chest pain, heamoptysis, lobar consolidation or collapse, pleural effusions, and pneumo- or hydropneumothorax (13, 27, 36, 41). In some cases expectoration of the larva has led to resolution of the symptoms. Most patients have had an accompanying eosinophilia, with a reported range of 30 to 72%, and when a pleural effusion has been present it has been eosinophilic in nature (41). Therefore, a triad of eosinophilia, subcutaneous swellings, and unexplained eosinophilic pleural effusion with a history of appropriate exposure risk should alert the physician to a diagnosis of gnathostomiasis (15).
Gastrointestinal manifestations are less common in humans but may present as sharp abdominal pains as the larva migrates through the liver and spleen or as a chronic mass in the right lower quadrant. Less commonly, there may be acute right iliac fossa pain with fever mimicking acute appendicitis or intestinal obstruction. Infection has also been found as an incidental (and asymptomatic) finding at surgery for a different problem. Radiologically the findings are of a thickened bowel wall with narrowing of the lumen (15). Histologically the lesions found at surgery are consistent with eosinophilic granulomas, but macroscopically they resemble neoplastic lesions, which has led to inappropriate radical surgery. This can be avoided only by consideration of possible parasitic infection in the preoperative work-up of such patients.
Involvement of the genitourinary tract is uncommon, but hematuria and the passage of the larva in the urine have been reported. Other symptoms attributed to Gnathostoma spp. include profuse vaginal bleeding, cervicitis, balanitis, an adnexal mass, and hematospermia (15, 36, 41).
The eye is the only organ in which the larva may be visualized, and therefore there are many more literature reports of ocular involvement than of involvement of other organs (15). Eye involvement has led to symptoms of uveitis (usually anterior), iritis, intraocular hemorrhage, glaucoma, retinal scarring, and detachment. The larva is usually be found in the anterior chamber and may be recovered intact, but there are a few reports of intravitreal localization (3, 4). The presence of eosinophilia is less likely in ocular disease because the eye is a privileged site, and it is usually mild if elevated at all (41).
Various reports have described a wide variety of manifestations, which include mastoiditis, sensorineural hearing loss, and extrusion of the larva from the external auditory canal, the soft palate, the cheek, the tip of tongue, and the tympanic membrane (15, 41).
Although gnathostomiasis was described in the 19th century, CNS involvement was proven only in the latter half of the 20th century, with the postmortem finding of a gnathostoma larva in the cervical cord of a patient with eosinophilic encephalomyelitis in 1967 (9). In the subsequent year the parasite was found on the surface of the cerebral hemisphere and attached to the choroid plexus of the lateral ventricle in two patients with fatal meningoencephalitis. There have been several case series of CNS disease, which has increased understanding of the pathophysiology (5, 38). Compared to other forms of disease, the CNS form of the infection carries the highest mortality, with reported rates of 8 to 25%, and 30% of survivors having long-term sequelae (5, 38, 41). However, these data are from the era before the use of albendazole and ivermectin.
The main features of CNS involvement are a radiculomyelitis, radiculomyeloencephalitis, eosinophilic meningitis, and subarachnoid hemorrhage. The hallmark symptoms are an acute onset of excruciating radicular pain and/or headache (subarachnoid hemorrhage or eosinophilic meningitis), with subsequent paralysis of the extremities and/or cranial nerve palsies. The typical clinical picture can be explained by the migratory pathway of the parasite, which gains entry to the spinal cord along nerve roots (cranial, cervical, thoracic, or lumbar), causing intense radicular pain (or headache in the case of cranial nerve or cervical root involvement) which usually lasts from 1 to 5 days (5). This initial pain is typically followed by a degree of paralysis ranging from weakness to complete paralysis of one to four limbs (the most common being paraplegia of lower limbs) as the parasite ascends the spinal cord to the brain. Urinary retention is usual with radiculomyelitis and radiculomyeloencephalitis. Cranial nerve palsies tend to occur after the onset of paralysis, and if multiple they signify a poor prognosis. Cerebral involvement is usually indicated by a depressed consciousness level or coma, but interestingly, mental confusion does not tend to occur (5, 38).
In some cases different CNS symptoms may occur, or current ones may reappear after a quiescent period due to further migration of the larva to another location within the CNS. This most commonly occurs within the first 2 weeks of the onset of the initial neurological deficit but may be seen as early as 6 h or as late as 1 month.
The pathogenicity in the CNS is thought to be the same as that elsewhere in the body, with direct mechanical injury causing the most damage due to tearing with or without destruction of the nerve tissue and its vascular structures, as well as inflammation and destruction of tissue due to toxin production. The hallmark signs of gnathostomiasis are hemorrhagic tracts, which have been well documented throughout the spinal cord and cerebral tissue postmortem (5, 24). Death occurs if vital structures in the brain stem are invaded, which may occur within 4 to 31 days following the onset of CNS symptoms, or if the larva burrows through a cerebral arteriole, resulting in massive subarachnoid hemorrhage (5). In Thailand 6% of subarachnoid hemorrhages in adults and 18% of those in children are due to gnathostomiasis (50).
The main differential diagnosis of CNS disease due to Gnathostoma is infection with Angiostrongylus catonensis, another highly prevalent parasite in southeast Asia. This may produce a similar eosinophilic meningoencephalitis, but the acute nerve root pain, signs of spinal cord compression, and hemorrhagic or xanthochromic spinal fluid seen in gnathostomiasis are absent with Angiostrongylus infection (38). The Gnathostoma larva is more invasive than that of Angiostrongylus and therefore produces more frequent focal neurological signs. In contrast, the Angiostrongylus larva, which is considerably smaller (120 μm wide and 12 mm long) and usually multiple, more commonly causes a meningoencephalitis, and although neurotropic, it is rarely fatal (21).
The triad of eosinophilia, migratory lesions, and obvious exposure risk are highly suggestive of the diagnosis of gnathostomiasis. Exposure risk must include residence in or travel to an area of endemicity and consumption of food that potentially contains the larval form of the parasite (raw or undercooked fish [in particular swamp eels, catfish, sleeper perch, bream, Nile tilapia, butterfish, loaches, or snake-headed fish], frogs, chickens, cats, or dogs). Clinically, the main differential diagnoses includes angiostrongyliasis, trichinosis, and cutaneous larva migrans.
Before the advent of serology, the diagnosis was more often made by the isolation of the larvae from the lesions they caused, but this is often difficult in migratory skin lesions, is clearly not practical for visceral disease, and is no longer required (Fig. (Fig.66).
Eosinophilia is frequently present during initial worm migration (as previously discussed), particularly in the skin or subcutaneous tissues, but is not always present at other times, and its absence should not exclude the diagnosis. Eosinophilia of the cerebrospinal fluid (CSF) is also highly supportive of CNS disease, with reported levels of 5 to 94% and a total CSF white cell count of up to 500/mm3 (range, 20 to 1420/mm3), but may also be found with several other parasites, e.g., Angiostrongylus cantonensis, Toxocara canis, Strongyloides stercoralis, Ascaris lumbricoides, Paragonimus westermani, Fasciola hepatica, and Trichinella spiralis and with schistosomiasis, neurocystercercosis, and other infections such as coccidiodomycosis and aspergillus infection (5, 21). Noninfectious conditions involving the CNS should also be considered (e.g., lymphoma, particularly Hodgkin's) (40).
In the 1960s skin tests using intradermal injection of G. spinigerum antigen were developed in Japan, but these were later shown to lack adequate sensitivity and specificity (26, 48). Scientists in Japan then went on to be the first to develop a serological test for the diagnosis, using crude somatic extract of adult Gnathostoma doloresi worms found locally (23). However, this was hampered by cross-reactivity with other locally found parasites, including Paragonimus westermani, Toxocara canis, Anisakis, and Fasciola hepatica. Later it was found that this was a problem particular to the G. doloresi antigen but could be overcome if the G. spinigerum antigen was used instead (1, 48).
Tests were improved by using antigen from the third-stage larvae (L3) of G. spinigerum instead of the adult worm, but cross-reactivity remained a problem with other parasitic infections. Furthermore, the L3 antigen was found to be highly complex, with over 20 components, some of which reacted with the sera of healthy controls (32). Tapchaisri and colleagues found that a specific L3 antigen with a molecular mass of 24 kDa had the greatest specificity and reacted only with gnathostomiasis sera and not with those from other parasitic infections (49). This subsequently formed the basis of the immunoblot and future trials on serodiagnosis (20).
An enzyme-linked immunosorbent assay (ELISA) for L3 immunoglobulin G (IgG) antibody was also developed, but the sensitivity was poor, ranging from 59 to 87%, with a specificity of 79 to 96% (1, 22, 46). Nuchprayoon and colleagues then showed that by using IgG subclasses rather than total IgG, the sensitivity and specificity of the ELISA could be improved. This had been shown previously to improve the diagnosis for other parasitic infections such as ascariasis, leishmaniasis, and filariasis (33). Their study showed that IgG2 had the least cross-reactivity (particularly with Angiostrongylus, the main differential diagnosis) and had a specificity of 88% and positive predictive value of 93%, whereas IgG1 had the highest sensitivity (98%) and negative predictive value (94%). With the exception of IgG2, cross-reactivity was found with Angiostrongylus cantonensis, hookworm, Strongyloides stercoralis, and Opisthorchis viverrini. They concluded that IgG1 antibody should be used as a screening test for those with presumptive gnathostomiasis and that IgG2 antibody could be used to confirm the diagnosis. A further extensive retrospective study of parasitic and nonparasitic diseases carried out over 7 years confirmed these findings of a lack of cross-reactivity (P. Dekumyoy, personal communication).
Currently a number of serological tests are available for the diagnosis of gnathostomiasis. The most widely used in Europe is an immunoblot to detect the specific 24-kDa band considered diagnostic of gnathostomiasis. This is carried out at various places, with the United Kingdom serology being sent to the Hospital for Tropical Diseases, Mahidol University, Bangkok, Thailand. The Swiss Tropical Institute (Basel, Switzerland), also performs serological testing. However, currently there are no commercial reagents available.
Specific mention should be made of the diagnosis of CNS disease. The CSF typically shows an eosinophilia, with elevated opening pressure and protein level. The use of CSF serology is not routine, and there are no published data on its utility in this context. Magnetic resonance imaging has also been useful in demonstrating the migratory lesions within the spinal cord (7, 43). The parasite produces high-signal intensity on T2 weighting and contrast enhancement on T1 weighting, which if static could be attributed to various vascular, neoplastic, inflammatory, or infectious causes. However, the migratory nature of such lesions would be consistent with the well-documented pathway of the Gnathostoma larva. Multiple sclerosis would also produce intermittent lesions on magnetic resonance imaging, but this can be excluded by the absence of CSF oligoclonal bands.
For many years there was no effective treatment for gnathostomiasis, and surgical excision of the larvae remained the only effective management. Various drugs were tried both in animal models and in humans without success, including thiabendazole, praziquantel, metronidazole, diethylcarbamazine, and quinine (18).
However, studies with albendazole in animal models were promising, and a trial by Kraivichian et al. in 1992 (n = 112) confirmed its efficacy in humans, with cure rates of >90% at a dose of 400 mg twice a day for 21 days (18, 30). Albendazole is a broad-spectrum benzmidazole antihelminthic which has proven efficacy against intestinal helminths and also extraintestinal helminthic infections such as opisthorchiasis, hydatid disease, and cutaneous larva migrans. The principal mode of action for albendazole is by its inhibitory effect on tubulin polymerization, with resultant loss of cytoplasmic microtubules. The loss of these impairs glucose uptake, with subsequent depletion of the organism's glycogen stores resulting in immobilization and death. Albendazole is poorly absorbed from the gastrointestinal tract (although it is absorbed better than mebendazole) and is rapidly converted in the liver to its active primary metabolite, albendazole sulfoxide, prior to reaching the systemic circulation. Oral bioavailability appears to be enhanced when albendazole is coadministered with a fatty meal (2). The drug has been found to be safe and relatively well tolerated, with the main side effects being nausea, dizziness, headache, and occasionally abnormal liver function tests and a transient leukopenia. Additionally, albendazole appears to stimulate the outward migration of the larva, thus making it more accessible and possibly amenable to excision (19, 47). The reason for this migration is unknown, but it tends to occur 7 to 14 days after commencing treatment.
Ivermectin has also been investigated for use in the treatment of gnathostomiasis, and it seems that its efficacy is similar to that reported for albendazole (30). Ivermectin binds selectively with high affinity to glutamate-gated chloride ion channels which occur in invertebrate nerve and muscle cells. This leads to an increase in the permeability of the cell membrane to chloride ions with hyperpolarization of the nerve or muscle cell, resulting in paralysis and death of the parasite (34). Ivermectin may also interact with other ligand-gated chloride channels, such as those gated by the neurotransmitter gamma-aminobutyric acid. It is metabolized in the liver and has few side effects, the main one being dizziness, and is therefore generally well tolerated. Ivermectin has been shown to be effective as either a stat dose of 0.2 mg/kg or as doses of 0.2 mg/kg on two consecutive days, and therefore adherence is not a problem (31). The studies involving ivermectin have had fairly small sample sizes (n = 17 to 21), and therefore its evaluation with a larger prospective study is needed. Of note is that a potential issue of flaring of disease was found in a study by Kraivichian et al., which showed that those in the ivermectin group experienced an exacerbation of cutaneous symptoms compared with those in the albendazole group (19). This could theoretically be a serious problem for visceral disease, in particular ocular and CNS disease, although no such occurrences have been reported. This raises the issue of steroid use, which for neurocystercercosis is used to prevent inflammation in sites which could potentially be hazardous. Theoretically this rationale for use could be applied to CNS or ocular gnathostomiasis; however, data are limited, and the only study to examine this (n = 162) showed no definite benefit in those who received oral or intravenous steroid treatment (38).
Initial treatment is not always successful, and second courses of treatment have been needed in some cases. Either albendazole or ivermectin may be used in sequence in such patients, and both have been used with successful outcomes for initial treatment failures (8, 25). However in regions of endemicity it may be difficult to differentiate between treatment failure and reinfection. A recent series reported by Strady et al., conducted with returning travelers outside areas of endemicity, indicated that initial relapse occurred up to 7 months after treatment, and the maximum period between two relapses was 15 months, with the latter occurring in one patient with CNS disease (45). A pragmatic approach would be that if the patient was asymptomatic after 12 months of treatment, this would be sufficient evidence of cure, particularly if supported by a resolution of any eosinophilia and a decrease in ELISA levels. In the event of relapse, a new 12-month follow-up period should be commenced.
Future trials would be useful to investigate the use of combined treatment with both albendazole and ivermectin and to determine whether relapse rates are lower with combination drugs than with monotherapy.
Eradication of the organism is unlikely given its global distribution and expanding culinary tastes for exotic foods in the West. Therefore, control will be achieved only through education campaigns to raise public awareness (22). Adequate cooking is the best way to ensure that the larvae are killed, although freezing infected meat to −20°C for 3 to 5 days is also effective. However, the common practice in Mexico of marinating fresh fish in lime juice is ineffective, and the organism has been found to be viable even after 5 days of immersion in lime juice (39). Public health education is essential to change the eating habits of people in areas with high levels of endemicity, and travelers need to be aware of the potential consequences of eating local delicacies. The large variety and wide distribution of animals that are intermediate hosts make it impossible to eliminate the parasite, and therefore appropriate cooking methods and avoidance of raw or undercooked freshwater fish are the only measures that can be taken. However, while travelers continue to seek the exotic and remote, gnathostomiasis will be seen with increasing frequency in the West and other regions where it is not endemic, and it should be considered another emerging imported disease of which physicians should be aware.
We thank Paron Dekumyoy for helpful discussion and Julie Watson and Patricia Lowe for use of photographic images.
Peter Chiodini is supported by the UCL Hospitals Comprehensive Biomedical Research Centre Infection Theme.
Joanna Herman is a specialist registrar in Infectious Diseases and Tropical Medicine at the Hospital for Tropical Diseases, London. She has an M.Sc. in Communicable Disease Epidemiology from The London School of Hygiene and Tropical Medicine. She has previously undertaken research in human immunodeficiency virus medicine. Her current research interest is in clinical parasitology.
Peter L. Chiodini read Zoology at King's College London and earned his Ph.D. in Parasitology at the Wellcome Research Laboratories. He then studied Medicine in London before undergoing specialist training in Communicable Diseases. He is Consultant Parasitologist at the Hospital for Tropical Diseases, Honorary Professor at the London School of Hygiene & Tropical Medicine, and Director of the UK Health Protection Agency (HPA) Malaria Reference Laboratory and the HPA Parasitology Reference Laboratory. His research interests include new diagnostic methods for parasitic infections, malaria, and hydatid disease.