57 Strongyloidosis

Strongyloidosis is an intestinal parasitism caused by the threadworm, Strongyloides stercoralis. The parasite, occurring in dogs, primates and man, is found throughout the moist tropics, as well as in temperate areas where poor sanitation or other factors facilitate the occurrence of faecally transmitted organisms. In some parts of the world, notably Africa and New Guinea, human infections caused by S. fülleborni have been reported (Hira et al. 1980). In Africa, the latter is primarily a parasite of primates, but in New Guinea, no animal host is known. S. stercoralis is unique among zoonotic nematodes, in that larvae passing in the faeces can give rise to a free-living generation of worms which, in turn, give rise to infective larvae. This life history alternative (i.e. heterogonic development) acts as an amplification mechanism, increasing the population of infective larvae in the external environment. The infective larvae are active skin penetrators; infection per os, while possible, is probably of limited importance. Because the parasitic female’s eggs hatch internally, a potential for autoinfection exists when precociously developing larvae attain infectivity while still in the host. This is another virtually unique feature of S. stercoralis infections in both its human and animal hosts. Autoinfection can occasionally escape control by the host, with massive re-penetration and larval migration. This can cause pulmonary or cerebro-spinal strongyloidosis as well as fulminant intestinal parasitism. Control of canine strongyloidosis has been achieved in kennels by strategic use of anthelmintics. Given the lack of epidemiological information community-based programs to control human strongyloidosis have not been attempted. The growing importance of human strongyloidosis depends upon the unique ability of S. stercoralis to replicate within its host and to behave as a potentially fatal opportunistic pathogen in immunocompromised hosts, particularly in those receiving corticosteroids.

The history of strongyloidosis and of its etiological agent, Strongyloides stercoralis, is presented in detail by Grove (1989). The disease, originally called Cochin China Diarrhoea, was discovered by the French naval physician Louis Normand in 1876. Meanwhile the parasite, known only on the basis of the rhabditiform pre-infective larvae passing in the faeces, was described by Bavay (1876), Normand’s colleague and professor of pharmacy at the naval hospital in Toulon. Bavay named the nematode Anguillula stercoralis and recognized that when the larvae were kept in faeces for a few days under favorable conditions, they developed into free-living adult male and female worms. Subsequently, in autopsies of soldiers returning from duty in Cochin China (presently Vietnam), Normand found larvae throughout the intestines, bile and pancreatic ducts and adult parasitic females in the intestines. Not surprisingly, given that the parasitic females (there are no parasitic males) differ markedly from the free-living adults in morphology, these parasitic females were considered a different species and named Anguillula intestinalis by Bavay (1877). Giving further credence to this deduction was the additional discovery of a second kind of larva, a filariform larva (subsequently recognized as the infective stage) which at that point in the history of strongyloidosis could logically be considered the larva of the putative second species, A. intestinalis. Thus, in the years immediately following the recognition of the disease and of the parasite, all the stages in the life of the parasite became known, but their relationship was confused because it appeared that there were two species with different life cycles, namely, A. stercoralis whose rhabditiform larva occurred in the stools and whose adults occurred in the external environment and A. intestinalis whose adults were intestinal parasites and whose progeny were filariform larvae.

Remarkably, Grassi and Parona (1878) almost immediately resolved some of the confusion and explained much of the unusual and complex life cycle of the parasite. They found that the parasitic female named A. intestinalis laid eggs which hatched rapidly, giving rise to the rhabditiform larvae that were known as A. stercoralis. Apparently, they had a homogonic strain of the parasite because all of the rhabditiform larvae developed to infective filariform larvae such as had been described for A. intestinalis. It remained for Perroncito in 1881 to complete the free-living part of the life cycle by showing that the rhabditiform larvae, as originally observed in faeces by Normand, do indeed develop into free-living males and females and, furthermore, that these in turn produce filariform larvae, the infective stage of the parasite. Perroncito, however, did not realize that the various life history stages he and others had observed were parts of a complex life cycle having facultatively alternating parasitic and free-living generations.

The French workers, Normand and Bavay, discovered the disease and described the parasite; the Italian workers, Perroncito and Grassi and Parona, elucidated the free-living life cycle; and, subsequently, the German parasitologists, Looss (1905), and Fülleborn (1914), found, respectively, that infection occurred by skin penetration, and that larvae could migrate from the skin to intestines via the circulation, lungs and trachea. Finally, Gage (1911) reported the occurrence of autoinfection, i.e. that larvae hatching from eggs laid in the host can develop to infectivity precociously and reinfect the same host in which they were hatched. The host, parasite and environmental factors that determine the alternative developmental pathways of S. stercoralis in both man and animals remain poorly understood but are under active investigation (see Life History below).

The family Strongyloididae (Class Secernentea, Order Rhabditida, Superfamily Rhabditoidea) is constituted of three genera, Strongyloides (Grassi 1879), Parastrongyloides (Morgan 1928), and Leipernema (Singh 1976). The members of the genus Strongyloides, also called threadworms, are heterogenetic, with both free-living and parasitic generations. The genus includes fifty-two named species. The majority of these are parasites of mammals, but some can be found in birds, reptiles and amphibians. The only species dealt with in detail in this chapter is Strongyloides stercoralis (Bavay 1876) (synonyms: Anguillula stercoralis, S. intestinalis, S. canis, S. felis) an intestinal parasite of dogs, primates and man. Strongyloides fülleborni (von Linstow 1905), is usually considered a parasite of primates that also infects humans, but, at least in some parts of its range, is transmitted in the absence of primates. Thus, its zoonotic status is presently uncertain and it is omitted here. Several species are important parasites of livestock (S. ransomi, S. westeri, S. papillosus) or are laboratory models of human strongyloidosis (S. ratti, S. venezuelensis). Parastrongyloides trichosuri, a parasite of Australian marsupials that is related to Strongyloides, is becoming an increasingly important laboratory model for some aspects of the Strongyloides life cycle, since it can be maintained in its free-living stages indefinitely as a microbiverous species on agar culture plates (Nolan et al. 2007).

The life history of S. stercoralis is complicated by a number of facultative alternatives. These alternatives include:

The latter pathway, when constrained by still poorly understood host and/or parasite factors, is thought to lead to a slow turnover in the adult worm populations, thus maintaining highly persistent chronic infection. When these constraints fail, autoinfection is explosive forming the basis for the fulminant hyper- and disseminated forms of the infection.

In its simplest homogonic form, the life cycle involves parthenogenetic females lying embedded in the crypts of the intestinal mucosa where they deposit their eggs. The egg hatches giving rise to a first-stage larva (L1), known as a rhabditiform larva in the parasitological literature. Just after hatching, the young L1 responds to environmental cues detected by 2 amphidial neurons to make the decision as to which pathway (homogonic vs heterogonic) by which it will develop (Ashton et al. 1998). If the homogonic route is selected, the actively feeding, microbiverous form leaves the crypts and moves down the intestines, and exits the body while still a pre-infective rhabditiform larva. During intestinal passage and in faecal deposits, this larva feeds, grows and moults so that two rhabditiform stages occur (L1, L2). Under favourable environmental conditions including a suitable faecal flora, warmth and moisture, the L2 grows and moults, giving rise to an infective filariform larva (L3), a long slender form, with a long slender oesophagus, hence the name filariform larva. It invades the host by active skin penetration.

In the heterogonic cycle, as in homogonic cycle, the larval progeny of the parthenogenetic parasitic females exit the host in the faeces. At this point (before the midpoint in the life span of the L1 (Nolan et al. 2004)) the larva again will use environmental cues sensed through its amphidial neurons to make a decision to follow the homogonic or heterogonic routes. For example at higher temperatures (above 34C) the worms tend to develop via the homogonic route (Nolan et al. 2004). If the heterogonic route is chosen the larvae will undergo 4 moults as they develop to the free-living adult male and female worms. These adults will mate and lay eggs, the larvae which hatch from the eggs will give rise to infective larva (L3) via the usual two rhabditiform stages. The infective forms leave the faeces or polluted soil in which they have developed and ascend surface particles to the extent permitted by soil or faecal moisture films (Sciacca et al. 2002). Here the larvae are positioned for contact with a host. The L3 will also respond to chemical (especially factors found in skin such as salt and urocanic acid (Forbes et al. 2003; Safer et al. 2007)) and heat gradients (Lopez et al. 2000) to move towards a host. After contact and percutaneous entry into a host, the larvae enter the circulation and migrate to the intestines. It is generally accepted that the migratory route involves the lung, trachea and oesophagus (pulmo-tracheal migration), but this route has been challenged by Wilson and by Schad and colleagues. After percutaneous infection of dogs, some larvae do follow the pulmonary route, but most do not. In fact, studies with radio labelled larvae have indicated that no predominant migratory route exists and that larvae reach the intestines via a number of different pathways (Mansfield et al. 1995).

Autoinfection is the third life history alternative. In this case, eggs hatch in the intestines as they normally do, but the larvae develop to infectivity precociously while still in the host. In this case the L1 not only choose the homogonic route of development, they also continue to develop as they move down the intestine, reaching the filariform L3 stage by the time they are in the colon. This autoinfective L3 (L3a) is morphologically distinct from the soil dwelling L3 (L3i) as it is smaller in size, slightly wider and has a less filariform oesophagus (Schad et al. 1993). The L3a penetrate the wall of the large intestine and from here migrate to many organs of the body, again, some use pulmonary migration, but others return to the intestine by other pathways. Upon return to the intestine, the L3a give rise to a fourth larval stage which in turn moults to give rise to the female adult worm.

The parasitic female is one of the stages found in tissue sections of the small intestine; it is rarely seen in the stools of infected hosts. It measures 2.0–2.8 mm in length and 37 m in width. It has a long cylindrical oesophagus, an intestine constituted of dorsal and ventral rows of 20 cells each. The tail is a short cone. The vulva is ventrally situated at 2/3 the body length from the anterior tip. Eggs are present in 2 single rows, one to either side of the vulva. The cuticle of the female worm is finely striated and, in tissue sections, is often wrinkled. In cross section, depending on the level, one may see a muscular oesophagus, intestine, ovaries and eggs. Reproduction is by parthenogenesis and, hence, there are no parasitic males.

The rhabditiform larva, the stage that hatches from the egg, is the form most commonly identified in faecal samples. It measures approximately 250 m in length and 17 m in diameter when passed in faeces, and is characterized by a bulbed oesophagus and a thinner, longer intestine (Schad 1989). In intestinal aspirates, the newly hatched larva is smaller, measuring 180–240 m in length and 14–15 m in width. In tissue sections they are often found in the intestinal submucosa and within small intestinal crypts, but only exceptionally in the lungs; they cannot be specifically identified based on their morphologic characteristics. The filariform (third stage) larva of autoinfective origin is the stage most frequently identified in parenteral tissues and body fluids (most often the sputum) in patients with disseminated infections. They are longer and more slender than rhabditiform larvae and have a cylindrical oesophagus that occupies one half the body length. In transverse sections the cuticle shows four characteristic lateral alae, which can be used for species identification. Filariform infective larvae arising as progeny of free-living adults are even longer and more slender, measuring up to 700 m in length and 20 m in width.

Chronic infections are probably sustained by a relatively low number of adult worms, many of which may be barren (the so-called post-reproductive females), which reside in relative harmony within their host’s intestine and the infection persists by means of periodic bouts of autoinfection (Schad et al. 1997). The occurrence and rate of autoinfection is generally believed to be regulated by the host’s cell-mediated immunity. When this regulatory function becomes impaired during immunosuppression, increasing numbers of autoinfective larvae complete the cycle, and the population of parasitic adult worms increases (hyperinfection). Eventually, with extraordinary numbers of larvae migrating, large numbers deviate from the generally presumed route (intestine—> venous bed—> lungs—> trachea—> intestine) and disseminate to other organs, including meningeal spaces and brain, liver, kidneys, lymph nodes, cutaneous and subcutaneous tissues. In these organs the larvae cause haemorrhage by breaking capillaries, elicit inflammatory responses, and implant Gram-negative bacteria carried from faecal material. The resulting syndrome, known as disseminated strongyloidosis, is nearly always fatal.

The validity of the migratory pathways in the above widely accepted model has been questioned by Schad et al. (1989) used an experimental canine model of disseminated strongyloidosis to show that only a few larvae could be recovered from the lungs of dogs with massive hyperinfection. Later, in studies based on the organ specific distribution of radio-labelled larvae and compartmental analysis of the data, they presented strong evidence that, in young dogs, the pulmonary route was not used by the majority of the migrating larvae. Larvae that began their migration in the skin (primary infection) or in the distal ileum (autoinfection) were not more likely to pass through the lungs than through any other organ, suggesting that the migratory pathway involved random dissemination throughout the body. However, this conclusion has not been fully accepted because large numbers of larvae are frequently identified in bronchoalveolar lavage fluid from hyperinfected human patients. It may be that in hyperinfection, as distinct from a primary infection, pulmonary migration is more frequent (Kerlin et al. 1995).

The theory that host immunity controls the rate and mode of parasite development fails to consider the role that parasites themselves may play in this regulation. It is known that the free-living nematode Caenorhabditis elegans can sense the population density and uses this information to switch developmental pathways. In both the dog and gerbil model of strongyloidosis initially low numbers of worms will lead to the autoinfective development of some of the larvae, an event not seen in infections initiated by higher numbers of worms (Nolan et al. 1999; Schad et al. 1997). The adverse impact of increased parasite density on egg production and growth (‘crowding effect’) has been demonstrated for several intestinal nematodes. Although it may be difficult to distinguish between host resistance and direct parasite-to-parasite effects, it seems clear that in a normal host-parasite relationship, the parasite may reach a particular population size or a critical biomass, after which yet unknown regulatory mechanisms intervene to limit the population. However, results of investigations using the gerbil-based model of strongyloidosis have failed to provide support for this type of self regulation of worm burden by the parasite (Nolan et al. 2002).

Genta (1986) has proposed that, during the parallel evolution of humans and their parasites, S. stercoralis developed the ability to reach an optimal population size in the duodenum of a human. If the initial infective dose of larvae is low, a higher rate of intraluminal moulting occurs, enabling the parasite to attain the infective stage internally, reinfect and multiply. This occurs until the ‘optimal’ size of the adult population is reached. In this model, it is assumed that S. stercoralis, similar to other nematodes, transmit their moulting signal by moulting hormones (ecdysteroids). As the size of the parasite population reaches a certain level, adult females decrease their production of ecdysteroids, resulting in a lowered moulting rate, i.e. just sufficient to replace the dying adults. During the initial phase of infection, the host mounts humoral and cellular immune responses directed at all tissue stages of the parasite. These well characterized responses do not eradicate all the parasites, but limit the size of the parasite population. Impaired immune responses may allow the growth of larger numbers of parasites, as reported in agammaglobulinemic patients, but total dysregulation of the parasite population does not occur since worms, in part, regulate their own growth. Conversely, the presence of intact immune responses is not sufficient to prevent dissemination should the parasites’ own regulatory mechanisms fail.

The level of ecdysteroid-like substances are generally negligible in healthy subjects. The administration of exogenous, or endogenous, corticosteroids may result in increased amounts of ecdysteroid-like substances in the host’s tissues, including in the intestinal wall, where adult females reside. These substances may act as moulting signals for the eggs or rhabditiform larvae, which transform intraluminally into excessive numbers of filariform larvae (Genta 1992). Available data are not sufficient to prove a dose-dependent effect, but it is indeed remarkable that patients who develop fulminating hyperinfection after only a few days of steroid administration are usually those who have received intravenous methylprednisone. Once a population has become very large (for example 100,000 adult worms) it may continue to expand rapidly, even at low moulting rates, and the discontinuation of steroids may not be sufficient to arrest the relentless growth process which leads to the host’s death. Some evidence from the gerbil model supports this as other immunosuppressive effectors are not as effective at inducing hyperinfection as is methyprednisolone and large worm burdens lead to hyperinfection in the absence of immunosuppression (Nolan et al. 2002).

Attempts to cultivate the parasitic stages of S. stercoralis in vitro have been unsuccessful. Infective larvae have been maintained for months under host-like conditions in tissue culture media (Chapman et al. 1994), but the larvae failed to grow or develop. Free-living stages are easily reared in standard parasitological coprocultures. These stages can also be raised on agar plates seeded with bacteria or in liquid cultures consisting of bacteria in a nematode saline.

These include dogs, various primates and humans. Cats, ferrets and gerbils (Nolan et al. 1993) have been infected experimentally. Transmissibility of S. stercoralis between host species varies geographically. The canine strain from North America is transmissible to humans (Georgi 1974), although molecular characterization using RFLP was able to distinguish a dog isolate from several human isolates (Ramachandran et al. 1997).

In both primates and dogs the pre-patent period is short, rhabditiform larvae appearing in the faeces in 1–2 weeks.

Symptoms and signs of infection vary markedly with respect to the individual. In primates there is also marked interspecific variation, monkeys being less susceptible than the anthropoid apes. Most cases in dogs are asymptomatic and become occult in 2–3 months. Although larvae disappear from the faeces, barren adult females may survive embedded in the intestinal mucosa for several months after the infection becomes inapparent (Schad et al. 1997). These infections can be reactivated by immunosuppression attributable to either chemotherapy or concurrent disease. Dogs that have expelled an infection are resistant to reinfection.

In young pups, hyperinfective strongyloidosis occurs spontaneously. Although these infections are usually mild and self-limiting, in some animals the worm burden may increase to clinically significant levels associated with watery or mucus diarrhoea and with signs of bronchopneumonia. Older dogs rarely become severely infected.

S. stercoralis occurs commonly in various monkeys. It is usually well tolerated, but in young Patas monkeys (Erythrocebus patas) it may produce severe hyperinfective strongyloidosis (Harper et al. 1982). These severely affected animals have diarrhoea, lose weight, and may die suddenly. Larvae may or may not be found in the faeces even in severely affected cases. Severe, often fatal strongyloidosis, occurs even more frequently in young anthropoid apes (Penner 1981). Gibbons are particularly susceptible to sudden death without a history of previous illness.

Diagnosis of S. stercoralis infection is complicated by the fact that larvae may be absent from the faeces even in symptomatic cases. Additionally, larvae (not eggs) pass in the faeces, making concentration techniques using high density flotation solutions somewhat difficult to use. The Baermann apparatus is commonly used for finding larvae in faeces. The first-stage larvae are easily recognized, their genital primordium being exceptionally prominent. Many cases of this infection are probably first suspected when larvae are seen either in a direct smear or in a saturated salt flotation. The Baermann funnel is then used to obtain clean, intact larvae for a definitive diagnosis. However, faecal flotations done with zinc sulfate yield readily identifiable larvae provided that the preparation is examined promptly before the larvae shrink. In animals showing respiratory symptoms, a transtracheal wash may reveal migrating third-stage larvae. This stage is easily identifiable by its long filariform oesophagus and its notched tail. A small percentage of the larvae present in a faecal sample may be third-stage larvae, particularly in recently acquired infections. Infectious stools held at room temperature for 24 hours or more may contain a variety of stages, including free-living adults. Although rarely used with dogs, the agar plate method of Koga et al. (1990) can also be used (for details see ‘Diagnosis’ under ‘Human hosts’ below). However, care must be taken to distinguish the species of the larvae leaving the tracks on the agar as hookworms are common in dogs and will also be detected by this technique. In the USA this parasite is most commonly diagnosed in puppies recently acquired from a pet store or ‘puppy mill’.

Intestinal pathology varies with intensity of the infection which in turn varies with the strain of organism and the age and species of host. In asymptomatic infected dogs, the intestinal tissues may be grossly normal and worms and larvae exceedingly difficult to find by histological methods. In symptomatic cases, gross intestinal changes range from congestion of mucosal surface with an abnormal abundance of luminal mucus to confluent ulceration that may penetrate to the muscular layer. In cases of severe infection, parasites in great abundance will be present in the intestinal walls (Grove et al. 1983).

In primates a similar range of lesions has been observed; however, severe strongyloidosis with significant ulcerative enteritis is rare in monkeys but well known in gibbons and orangutans. Complicated strongyloidosis with severe hyperinfection occurs spontaneously in young Patas monkeys, gibbons and orangutans. Characteristic pathological lesions include the presence of the full spectrum of parasite life history stages in the gastrointestinal tract and the presence of filariform larvae in the lungs associated with pulmonary haemorrhage. The number of migrating larvae frequently does not correlate with the amount of pulmonary haemorrhage.

Treatment of dogs with an active hyperinfection is difficult because available drugs do not kill the migrating autoinfective L3. However, unless a dog is very young or immunosuppressed, it is unlikely to have numerous migrating autoinfective larvae at any one time. The following anthelmintic treatments will remove adult S. stercoralis from dogs:

In all cases, follow-up faecal examinations should be done weekly for 2 to 3 weeks to verify that no migrating larvae survived the treatment and matured. In cases where hyperinfection is suspected the following treatments can be used: fenbendazole, once daily for 7 to 14 days at 50 mg/kg or ivermectin once every 4 days for 3 or 4 doses at 200 ug/kg (Mansfield et al. 1992). Although these treatments will not kill migrating larvae, they will remove adults as they mature in the small intestine and therefore prevent new autoinfective larvae from being produced. The problem is that the life-span of migrating autoinfective larvae is unknown (although mathematical modelling suggests the migration should take no more than 5 days (Mansfield et al. 1995)), and, therefore, recommended treatments may continue for longer than necessary. Again, follow-up faecal examinations should be done to confirm that a parasitological cure has been achieved. Ivermectin and fenbendazole should also be effective against S. stercoralis infections in cats.

The possible occurrence of migrating autoinfective larvae must also be considered in treating primates for a S. stercoralis infection. In most animals, unless immunosuppressed, it is unlikely that numerous autoinfective larvae will be present, and, therefore, a single course of treatment should be curative. The following treatment regimes, each repeated after 2 weeks, have been used in primates:

However, in the anthropoid apes, i.e. gibbons, chimpanzees, gorillas, and the orangutan, fatal infections have occurred, especially in juveniles in the absence of immunosuppression. Fatal hyperinfections are also seen in otherwise normal Patas monkeys. Thus, in these animals a more extended course of treatment may be advisable. In all cases, follow-up faecal examinations should be done for an extended period to verify that treatment has completely eliminated the parasites.

The prognosis for dogs infected with S. stercoralis is good. Except in dogs infected with the Southeast Asian (Indochinese) strain of the parasite, the infection is usually self-limiting and infrequently attains a clinical level of intensity. The prognosis for infections in primates varies with the species of host. Most monkeys carry easily treated asymptomatic infections. Anthropoid apes, on the other hand, are more susceptible to severe strongyloidosis, and young gibbons and orangutans, in particular, may die suddenly without apparent previous illness. In anthropoid apes, partial clearing of the infection by most anthelmintics occurs, resulting in low-grade, sometimes occult, chronic infection which may be seriously exacerbated by subsequent immunosuppression.

The pre-patent period in humans has been reported to be between 23 and 28 days (Freedman 1991). This is about a week longer than the pre-patent period in animals.

The clinical manifestations of acute strongyloidosis are generally associated with the path of larval migration to the small intestine. Infected individuals may experience irritation at the site of skin penetration by larvae, followed by tracheal irritation or dry cough and ultimately gastrointestinal symptoms such as diarrhoea, constipation, abdominal pain, or anorexia (Keiser 2004).

In uncomplicated strongyloidosis, many patients are asymptomatic or have mild cutaneous and/or abdominal symptoms.

The gastrointestinal manifestations of chronic strongyloidosis are usually non-specific. Epigastric abdominal pain, post-prandial fullness or bloating, and heartburn are among the symptoms most commonly reported, episodes of diarrhoea alternating with constipation may also occur (Grove 1996; Milder et al. 1981). The diarrhoea usually consists of semi-formed non-bloody stools. Occult blood in the stool can occur in persons with chronic infections, and even massive colonic haemorrhage has been reported.

Physical examination of chronically infected patients is normal or reveals only mild abdominal tenderness on palpation. Less commonly, chronic strongyloidosis resembles inflammatory bowel disease, particularly ulcerative colitis, and the endoscopic appearance may be that of pseudopolyposis.

Dermatologic manifestations such as recurrent urticaria can occur (Leighton 1990; Pelletier et al. 1988) as can larva currens, (pruritic linear streaks located along the lower trunk, thighs and buttocks) as a result of migrating larvae.

Unusual manifestations of chronic strongyloidosis include arthritis (Richter et al. 2006), nephrotic syndrome (Hsieh et al. 2006), chronic malabsorption (Alam 1982; Atul et al. 2005; Garcia et al. 1977; Harish et al. 2005; Sturchler 1987), duodenal obstruction (Suvarna et al. 2005), focal hepatic lesions (Gulbas et al. 2004) and recurrent asthma (Tullis 1970).

Hyperinfection describes the syndrome of accelerated autoinfection, generally the result of an alteration in immune status (Longworth et al. 1986). The distinction between autoinfection and hyperinfection is not strictly defined, but hyperinfection syndrome implies the presence of signs and symptoms attributable to increased larval migration. Development or exacerbation of gastrointestinal and pulmonary symptoms is seen, and the detection of increased numbers of larvae in stool and/or sputum is the hallmark of hyperinfection. Disseminated infection occurs when larvae migrate beyond the organs of the autoinfective cycle (lung and gastrointestinal tract), although this may occur at low levels in chronic strongyloidosis.

Hyperinfection syndrome has been described as late as 64 years after an individual has left a Strongyloides-endemic area.

Gastrointestinal symptoms commonly occur and may include crampy abdominal pain or bloating, watery diarrhoea, constipation, anorexia, weight loss, difficulty swallowing, sore throat, nausea or vomiting. Diffuse abdominal tenderness and hypoactive bowel sounds may be due to ileus and small bowel obstruction. Protein losing enteropathy can give rise to hypoalbuminemia with peripheral oedema and ascites. Mesenteric lymphadenopathy has been reported to cause intestinal pseudo-obstruction in HIV infected patients with hyperinfection syndrome. Mucosal ulceration can occur in the small intestine as a result of direct invasion of larvae and may be associated with occult blood, haematochezia or life threatening gastrointestinal bleeding.

Penetration of large numbers of larvae through the intestinal wall can be associated with gram negative sepsis as larvae carry organisms with them into the bloodstream. Organisms that have been reported to cause sepsis in such patients include Group D streptococci, Streptococcus bovis meningitis and bacteremia, Escherichia coli, Klebsiella pneumoniae, Proteus mirabilis, Pseudomonas, Enterococcus faecalis, coagulase negative staphylococci and Streptococcus pneumonia (Keiser 2004).

Either aseptic or gram negative meningitis can be associated with disseminated strongyloidosis. Larvae have been recovered from cerebrospinal fluid, meningeal vessels, dura, epidural, subdural and subarachnoid spaces.

Pulmonary manifestations, if present, include cough, wheezing, hoarseness, palpitations, atrial fibrillation, pleuritic chest pain, or dyspnea. Petechial haemorrhage, hyperemia of the bronchial mucosa or, rarely, massive haemoptysis have been reported. Chest radiographs most frequently demonstrate focal or bilateral interstitial infiltrates.

Cutaneous manifestations such as, Cutaneous periumbilical purpura has been described in patients with disseminated disease due to migration of larvae through vessel walls in the dermis (Salluh et al. 2005).

Apart from the meningitis (both bacterial and aseptic), a less common form of central nervous system manifestation is the formation of cerebral and cerebellar abscesses containing S. stercoralis larvae.

Infrequent manifestations such as cardiac arrhythmias and arrest are rare and have been attributed to a direct myocardial damage caused by the migrating larvae or to electrolyte imbalance precipitated by severe intestinal strongyloidosis. The passage of larvae in the sperm and the presence of genital lesions in association with strongyloidosis have also been described, however, Grove (1982) found no evidence that it could be transmitted through sexual contact.

The stage of S. stercoralis most commonly identified in faeces is the rhabditiform larva, but filariform larvae, adult females and even eggs also may be identified. The sensitivity of a single stool examination for the detection of S. stercoralis ranges between 30% and 60%. The use of a Baermann apparatus allows a larger volume of faeces (up to several grams) to be examined and is more sensitive than direct microscopy. Culturing faeces mixed with bone charcoal or peat moss also increases the sensitivity of faecal examination. However, these procedures are not suited for routine diagnosis in the clinical laboratory. A detection method proposed by Koga et al. (1990) involves the use of nutrient agar plates and is very sensitive and easy to perform. This method depends on larval dispersal from a faecal sample applied to the surface of a 1.5% nutrient agar plate and the subsequent growth of bacteria along the tracks left by these larvae. This technique has been shown to be more sensitive than formalin-ether sedimentation, Harada-Mori filter paper cultures or the Baermann technique (Arakaki et al. 1990; Intapan et al. 2005; De Kaminsky 1993), and using a cheaper non-nutrient agar works almost as well (Sukhavat et al. 1994). Although the examination of duodenal aspirate is reportedly very sensitive, this invasive method is recommended only in the paediatric patient when it is necessary to achieve a rapid demonstration of parasites, as in an immunocompromised child with suspected overwhelming infection. The ‘string test,’ a gelatin capsule containing a string swallowed by the patient and retrieved after a few hours, enjoyed some popularity a few years ago, but currently is infrequently used. In disseminated infections larvae of all stages and adult parasites have been found in specimens of sputum and broncho-alveolar lavage, ascitic fluid, pancreatic aspirates, and cerebro-spinal fluid. In summary, stool examination is currently the primary technique for the detection of S. stercoralis. If special techniques are not available, several specimens collected on different days should be examined if the diagnosis is strongly suspected.

The only haematologic abnormality associated with chronic, uncomplicated strongyloidosis is eosinophilia. In most case series, Strongyloides-associated eosinophil levels ranged between 500 and 1,500 eosinophils/mm³. Patients with disseminated strongyloidosis often have normal eosinophil counts likely related to either corticosteroid administration or endogenous factors (e.g. fever, endogenous steroids, stress, epinephrine).

Total serum IgE levels are elevated (> 200 IU/ml) in 50 to 70% of the patients with strongyloidosis. The diagnostic relevance of such elevation is similar to that of eosinophilia. Both eosinophilia and an elevated IgE level should be investigated, but the absence of elevation does not necessarily exclude strongyloidosis.

Enzyme linked immunosorbent assay (ELISA) has been increasingly used in conjunction with stool studies to increase diagnostic sensitivity. The high negative predictive value of the ELISA can be particularly useful in excluding strongyloidosis part of the differential diagnosis (Genta 1988). Despite its usefulness, serodiagnosis has several limitations including cross reactivity in patients with active filarial infections, lower sensitivity in patients with haematologic malignancies or HTLV-1 infection and inability to distinguish between current and past infection (Conway et al. 1993a, b). In addition, the current ELISA relies on the preparation of larval antigen from stool samples of heavily infected humans or experimentally infected animals.

Various techniques have been developed in an effort to improve on the drawbacks of the current immunoassays. Recombinant antigens, such as NIE, have been proposed as an alternative to the crude antigen currently in use (Ravi et al. 2002). An immediate hypersensitivity skin test has been used in a research setting quite effectively, although it may have limited utility in HTLV-1 infected patients (Neva et al. 2001). Most recently, a lucifease immunoprecipitation assays (LIPS) has been developed using two recombinant Strongyloides-specific recombinant construct that has provided a rapid, sensitive and specific method for diagnosis (Ramanathan et al. 2008).

The pathologic lesions associated with chronic, uncomplicated S. stercoralis have received little attention, because only rarely have patients with such lesions come to autopsy. However, pathologic descriptions of the lesions in a few patients in whom strongyloidosis was an incidental finding and animal studies indicate that the worms can exist in the intestinal mucosa without causing significant inflammatory responses or tissue damage. The classic description of the pathology of strongyloidosis was made by De Paola in 1962, and later updated by Genta and Caymmi-Gomes. These authors proposed the subdivision of the intestinal lesions into three distinct forms.

In ‘catarrhal enteritis’ (presumably associated with light infections), the small intestine is congested, the mucosa is covered with abundant mucoid secretions, and scattered petechial haemorrhages are present. The most remarkable histologic feature is an increased mononuclear infiltrate in the submucosa, although parasites are rare. In the more severe ‘oedematous enteritis’, the intestinal wall is grossly thickened, the mucosal folds flattened, and the affected intestinal segments assume a rubbery consistency. Submucosal oedema, flattening of the villi, and parasites scattered throughout the lamina propria are observed microscopically. The most severe form ‘ulcerative enteritis’ is almost exclusively seen in association with hyperinfection. The intestinal walls may be rigid due to the oedema and fibrosis resulting from long-standing inflammation, the mucosa may be show atrophy, erosions and ulcerations. An abundant inflammatory infiltrate, most often consisting of neutrophils, as well as all stages of S. stercoralis, are present throughout the intestinal mucosa. Jejunal perforation has been reported in patients with the ulcerative enteritis form of strongyloidosis. Uncommonly, the mucosal damage occurs predominantly in the large intestine, simulating ulcerative colitis and pseudopolyposis. S. stercoralis larvae have been found in the appendix, and eosinophilic appendicitis apparently caused by this parasite has been reported. In patients with disseminated strongyloidosis, the intestinal lesions reflect the large number of worms dwelling within the small intestinal mucosa and penetrating the intestinal walls. In addition, the stomach and the peritoneal cavity may be invaded by migrating parasites. However, because most of these patients are receiving immunosuppressive doses of corticosteroids, inflammatory responses are often minimal in spite of extensive tissue damage. The gastrointestinal pathology is often overshadowed by the lesions found in other organs, particularly in patients who receive anthelminthic therapy before succumbing to disseminated strongyloidosis.

Migrating parasites may cause mechanical damage as well as inflammation. In human patients, the extra-intestinal organ most commonly affected by this migratory damage is the lung. In severe disseminated infection, when larger numbers of adult parasites dwell in the intestine and millions of larvae migrate throughout the body, alveolar microhaemorrhages may result in massive pulmonary bleeding. As larvae penetrate the large intestine, they create small breaks in the mucosa that facilitate the invasion of the bloodstream by enteric bacteria. The larvae themselves carry bacteria on their cuticle to distant sites. Regardless of the mechanism, the widespread dissemination of larvae is frequently associated with polymicrobial sepsis, diffuse or patchy bronchopneumonia, pulmonary and cerebral abscesses, and meningitis. Filariform larvae, and occasionally rhabditiform larvae and adult worms, also may disseminate to mesenteric lymph nodes, the biliary tract, as well as the liver, pancreas, spleen, heart, endocrine glands and ovaries. In these locations the parasite frequently induces a granulomatous response.

To prevent the development of hyperinfection syndrome, chronically infected, asymptomatic individuals must be treated (Coulter et al. 1992). Because even one remaining adult female can multiply and cause disseminated disease, the goal of treatment is complete eradication of the parasite. The current treatment of choice for chronic strongyloidosis is single dose ivermectin (200ug/kg), although some studies have suggested that 2 doses of ivermectin 200 ug/kg given on consecutive days may have greater efficacy. Ivermectin has better efficacy when compared to thiabendazole in patients with chronic, uncomplicated strongyloidosis. In a randomized trial, up to 95% of patients on thiabendazole experienced side effects compared to 18% of ivermectin-treated patients. Side effects of thiabendazole include general fatigue, dizziness, headache, nausea, anorexia, abdominal pain, liver dysfunction and neuropsychiatric symptoms. Ivermectin has superior efficacy when compared to albendazole, a drug that has cure rates ranging from 45–77% (Archibald et al. 1993; Jorgensen et al. 1996).

For disseminated strongyloidosis, oral ivermectin should be given daily until stool examinations are negative for at least 2 weeks (the duration of the autoinfective cycle) or longer if patients remain immunosuppressed. Off label rectal administration of ivermectin or thiabendazole, while useful in some critically ill patients, can be problematic in patients with severe diarrhoea. Patients with paralytic ileus can have difficulty absorbing oral ivermectin due to tissue oedema, larger volume of distribution and increased clearance of unbound drug. Lower serum ivermectin levels than that achieved in normal subjects after oral administration have been demonstrated in patients with paralytic ileus. Parenteral formulations of ivermectin, however, are not currently approved for use in humans and have only been used under compassionate use INDs.

Except in cases of hyperinfection, the prognosis is good. Many infected persons are asymptomatic or have nonspecific minor complaints. In adults, these well-regulated infections may persist for decades without producing clinically significant strongyloidosis, and most will respond to anthelmintic treatment. However, because the risk of developing severe, hyperinfective strongyloidosis is always present, all infected persons must be considered at risk for fatal infection and treated with the goal of achieving a parasitological cure. Once hyperinfection occurs prognosis should be guarded.

Although information regarding the worldwide prevalence of strongyloidosis is fragmentary, 3 million to 100 million are estimated to be infected worldwide. The unreliability of these estimates is reflected in the wide range of prevalence rates, varying between <1% and 85%, of populations living in adjacent regions of the same country. With these limitations in mind, one can assume that S. stercoralis is present in virtually all tropical and subtropical regions of the world. Pockets of low endemicity (<1% to 3%) exist in several industrialized countries of Western Europe (e.g. Italy, France, and Switzerland), Eastern Europe, the USA (the Appalachian region and the Southern states), Japan (Okinawa) and Australia (aboriginal populations). Significant prevalences of strongyloidosis have been found in institutionalized patients. Considering the long persistence of this parasite in its host and its relatively high prevalence among some populations, physicians practicing in industrialized countries should consider strongyloidosis in immigrant or refugee patients born in tropical or subtropical regions as well as in persons from local areas of endemicity (Genta 1989).

For reasons that are not entirely clear, corticosteroids have a particularly strong and specific association with the development of hyperinfection syndrome. Hyperinfection syndrome has been described regardless of dose, duration or route of administration. Even short courses of steroids in immunocompetent patients have led to hyperinfection syndrome and death. Other therapies or conditions may predispose to dissemination although the concomitant administration of steroids in most cases makes it difficult to assign a direct causal association (Ramanathan et al. 2008).

A growing body of evidence points to the synergistic relationship between human T-cell lymphotropic virus type 1 (HTLV-1) and Strongyloides. Higher rates of Strongyloides infection have been found in HTLV-1/Strongyloides-co-infected patients (Arakaki et al. 1992). Relapsing Strongyloides infection despite treatment should prompt consideration of HTLV-1 infection. HTLV-1 enhances susceptibility to Strongyloides infection as a result of diminished IgE levels. Strongyloides may, in turn, facilitate HTLV-1 virus replication as suggested by a measurable decline in HTLV-1 mRNA levels in one patient after treatment with ivermectin. Strongyloides has been proposed to accelerate the progression of HTLV-1 to adult T-cell leukemia in that there is a more rapid development of leukemia in co-infected patients.

Hyperinfection syndrome has not been observed frequently with HIV infected patients despite vast numbers of co-infected individuals. A recent study postulates that lower CD4+ counts may favour indirect rather than direct development of Strongyloides larvae based on the proportion of free living adults and infective larvae in stools of co-infected patients. Whether immune reconstitution syndrome occurs after the initiation of antiretroviral therapy in Strongyloides infected patients remains unclear although this issue has been raised in case reports.

Several case reports have supported an association between Strongyloides infection and primary hypogammaglobulinemia. In these cases, prolonged recovery and refractoriness to therapy was noted. Haematologic malignancies such as lymphoma have been associated with hyperinfection syndrome in the absence of corticosteroid use. Relatively few cases of infection following bone marrow transplantation have been reported.

Transmission of S. stercoralis among both humans and animals can be prevented by implementing measures aimed at ensuring proper disposal and treatment of excrement and by avoiding contact with contaminated substrata, i.e. soil, caging, etc. The free-living larval stages are susceptible to desiccation. Thus, maintaining a clean dry environment provides effective control. In human patients from endemic areas who may harbour asymptomatic chronic strongyloidosis, life-threatening disseminated hyperinfection may be prevented by seeking and eradicating the parasite before corticosteroid, immunosuppressive or anti-neoplastic therapy is started. Strongyloidosis in animal populations has been a problem in canine breeding kennels and in the primate colonies of zoos and research organizations. Both death losses and occurrences of clinical strongyloidosis can be reduced in breeding kennels by periodic mass treatment with thiabendazole or ivermectin, but this will not eradicate the parasitism in the dog population. Apparently, eradication has been achieved as a by-product of Filaroides hirthi control. This involved treating of brood bitches between pregnancies with albendazole given at the rate of 25 mg/kg orally twice daily for 5 days. Control of strongyloidosis in primate colonies has depended on creating a clean dry environment because the free-living larvae of strongyloides are highly susceptible to desiccation. It also depends on the detection of infected individuals by faecal examination and their treatment. Thiabendazole given orally by stomach tube or in food as a single dose of 100 mg/kg and repeated after 2 weeks is effective for control.