Necatoriasis: treatment and developmental therapeutics.

Two hookworm parasites, Necator americanus and Ancylostoma duodenale, infect approximately one billion people worldwide. These hookworms are one of the leading causes of iron-deficiency anaemia especially in children, resulting directly from intestinal capillary blood loss following the feeding activities of fourth-stage (L4) larva and adult worms. If ignored, human hookworm infections can retard growth and the intellectual development of children. Another clinical manifestation often associated with hookworm infections is cutaneous larva migrans (CLM). It is a well recognised, usually self-limiting condition caused by the infectious larvae of nematodes, especially Ancylostoma spp. CLM is characterised by skin eruption and represents a clinical description rather than a definitive diagnosis. Of the hookworm parasites, the dog and cat worm A. braziliense and A. caninum are the most common nematodes causing CLM, although many other species have also been implicated. The major subject of this review article will be discussion of the evolution of therapies and treatment of human necatoriasis and the development of experimental infections with N. americanus. Difference in the clinical efficacy of mebendazole and albendazole will be discussed along with drug resistance of N. americanus.


Introduction
N. americanus and A. duodenale are the causative agents of necatoriasis and ancylostomiasis, two hookworm infections common in man. Together with ascariasis, these hookworm parasites remain the most common intestinal nematodes in the world with significant economic, social and medical impact [1-3].
N. americanus, a hookworm parasite of the family Ancylostomatidae, subfamily Necatorinae, is distinguished by the presence of two chitinous cutting plates in the buccal cavity and fused male copulatory spicules. Adult N. americanus parasites attached to the villi of the small intestines will suck blood. The infective larvae of N. americanus, by being obligatory penetrators, secrete all mechanistic classes of proteolytic enzymes having two overall pH optima of 6.5 and 8.5 [4]. Hookworm larval secretions were shown to degrade a host of human skin macromolecules including collagen types I, III, IV and V, fibronectin, laminin and elastin [4,5]. With the exception of collagen Type V, all other skin macromolecules tested were hydrolysed by the activity of the N. americanus aspartyl proteinase. The latter, in turn, is inhibited by pepstatin A [4].
Utilising a number of different but complimentary approaches, Brown and Pritchard [6] have demonstrated the immunogenicity of the N. americanus acetylcholinesterase in infected individuals. Acetylcholinesterase is secreted by wide range of parasitic nematodes including N. americanus [7,8]. It plays a potentially parasite-protective role [9] due to the presence of a conserve amino acid sequence in the active site regions of the cholinesterases [10] which allows for human acetylcholine to be a suitable substrate for the parasitic acetylcholinesterase.
Using adult N. americanus originating from Togo (Africa) and Sarawak (Malaysia), Romstad et al. [11] have found differences in the length of the nucleotide sequences of the second internal transcribed spacer of parasitic rDNA, suggesting that there is either population variation in the sequence of N. americanus, or the more likely conclusion is that N. americanus from the two countries may have represented genetically distinct but morphologically similar (i.e., cryptic) species.
In general, if lesser parasitic burden is present, the resulting necatoriasis is asymptomatic. The same may also be the case during larval pulmonary migration. However, symptomatic pulmonary disease may occur as consequence of conditions such as Löffler's syndrome, the effects of larval tissue migration, airway reactivity or bronchospasm, infectious bacterial complications from parasitic migration and associated aspiration and rarely, from chronic eosinophilic pneumonia, transdiaphragmatic penetration, or symptoms of upper airway obstruction. It is manifested by fever, cough, chest pain, haemoptysis, dyspnea and wheezing [2]. The most serious clinical manifestation associated with human necatoriasis is iron-deficiency anaemia [12]. To this end, the intestinal blood loss caused by hookworms is proportional to the number of adult parasites in the gut. Even though, as compared to ancylostomiasis, anaemia resulting from necatoriasis is generally considered to be less severe, there has been no preponderance of evidence to suggest that endemic A. duodenale infection has had greater impact than N. americanus infection on the iron status of the affected populations [13].
Between the asymptomatic state and the presence of iron-deficient anaemia, patients with necatoriasis may display a host of symptoms that include • cutaneous manifestations of pruritis and dermatitis (commonly referred to as 'ground itch') which develop after penetration of infective larvae through the skin • pulmonary manifestations such as cough, wheezing, bronchitis, or pneumonitis which occur during larval migration through the pulmonary tree to the gastrointestinal tract • gastrointestinal manifestations such as abdominal discomfort often with postprandial accentuation, nausea, vomiting, diarrhoea (usually with melena) and cramps, anorexia and the loss of weight bifurcum (a nodular worm) is of major health concern in northern Togo and Ghana, where N. americanus also exists at high prevalence.

Cross-reactivity between N. americanus, Schistosoma mansoni and Ascaris lumbricoides
Polyparasitism is commonly observed in endemic communities and reactivity of sera from hookworminfected patients against schistosomular and ascaris antigens has also been reported. The protective cross-immunity between N. americanus and Schistosoma mansoni has been investigated in NIH and BALB/c mice and its implication with respect to sero-diagnosis has been discussed by Timothy et al.
[37]. While protective resistance to homologous challenge with both parasites has been confirmed, functional immunity to heterologous challenge was not demonstrated. Furthermore, sera from mice that had received homologous challenges with N. americanus and from hookworm-infected animals, which had previously been exposed to radiationattenuated S. mansoni, exhibited an enhanced IgGA/IgGM response to infective stage N. americanus somatic antigens.
After studying a population from Papua New Guinea infected predominantly with N. americanus, Pritchard et al.
[38] have presented sero-epidemiological evidence demonstrating the presence of a high degree of cross-reactivity between IgG, IgA and IgM antibody responses to N. americanus ES and Ascaris lumbricoides PCF antigens. The results suggested that the cross-reactivity was due to epitopes carried on a range of Ascaris antigens and that a number of Necator-specific antigens do exist. The observed cross-reactivity accounted for a peak in the antibody levels against N. americanus in 10 -13 years old children (driven by infection with A. lumbricoides) as well as in the maintenance of apparent antibody levels against A. lumbricoides in older age groups (driven by infection with N. americanus in the absence of overt infection with A. lumbricoides).

Evolution of therapies and treatment of necatoriasis
The potent antiparasitic activity of two benzimidazole drugs, mebendazole (5-benzoyl-2-benzimid azolecarbamic acid methyl ester) and albendazole (methyl 5-propylthio-2-benzimidazolecarbamate) has been used very effectively against human hookworm infections [39-48]. The relatively low cost of mebendazole [49] has been an important factor to consider, especially in the treatment of hookworm infestations in impoverished communities.
Initial studies indicated that albendazole may have elicited its antihelminthic activity by selectively blocking the glucose uptake by adult worms lodged in the intestine and their tissue-dwelling larvae [50]. Inhibition of the glucose uptake is thought to result in the endogenous depletion of glycogen stored within the parasite thereby causing a decrease in the formation of adenosine triphosphate, which is essential for the reproduction and survival of the parasite. Recent developments, however, have demonstrated that benzimidazole anthelminthics may exert their antiparasitic activity via binding to the nematode β-tubulin. By binding to free β-tubulin, benzimidazoles inhibit the polymerisation of tubulin and the microtubule-dependent uptake of glucose [51,52]. This mechanism of action is believed to account for the association between tubulin alleles and drug resistance.
Albendazole appeared to be more effective than When used against hookworm infections, a standard regimen of pyrantel pamoate is applied usually as a single oral dose of 11 mg/kg, to a maximum dose of 1.0 g [35].
Ivermectin, the 22,23-dihydro analogue of the macrocyclic lactone avermectin B 1 , has displayed antiparasitic activity in vitro [77] and against an array of nematode and arthropod parasites in veterinary medicine [78][79][80]. When administered as a suspension into to the stomach of hamsters by oral intubation, ivermectin at 30 mg/kg completely cleared pre-adult N. americanus; a regimen consisting of 10 mg/kg (followed by a repeated dose on day later) achieved the same result [81]. However, hamsters carrying adult N. americanus worms were completely cured of infection by doses of 15 mg/kg (single dose) and 7.5 mg/kg (repeated on the next day).
Xia et al. [82] studied the clinical efficacy of ivermectin in the treatment of several human nematode infections. As compared to ascariasis and trichuriasis, the therapeutic efficacy of ivermectin against hookworm infections (N. americanus and A. duodenale) was rather disappointing. Thus, when tested at single oral doses of 0.1 and 0.2 mg/kg, the cure rates against ascariasis, hookworm and trichuris infections were 100%, 3.8% and 50% and 95.5%, 11.8% and 76.5%, respectively. By comparison, pyrantel pamoate at a single oral dose of 10 mg/kg showed cure rates of 95.5%, 29.6% and 31.6%, respectively. The observed side effects were mild and transient in all groups [82]. Naquira et al. [83] also reported little therapeutic efficacy of ivermectin in patients with hookworm infections (A. duodenale, N. americanus) following medication with 50, 100, 150, or 200 µg/kg as a single dose, or a two-day course with either 100 or 200 µg/kg of the drug. Similarly, Whitworth et al. [84] reported no apparent effect of ivermectin on N.
americanus infection in man when tested at 150 mg/kg in a double-blind placebo-controlled trial conducted in Sierra Leone.
The effects of a new anthelminthic drug, tribendimidin {N,N-[bis-4'-(1-dimethylaminoethylidene amino)phenyl]-1,4-phenylenedimethylidyne amine} on the cuticle of N. americanus was evaluated in golden hamsters [85]. One hour after medication with a single dose of 150 mg/kg, some worms showed cuticular swelling, fusion of transverse striations and attachment of host leukocytes on the parasite's damaged cuticular sur face. At four hours post-treatment, the cuticle revealed moderate swelling or even erosion; the ventral cutting plates appeared to be swollen. After 8 -24 hrs, severe cuticular swelling, erosion and peeling in female worm tails and male copulatory bursa, were observed. After 4 -8 hrs post-treatment, there has been no increase in lesions in the small intestinal mucosa of the hamsters. The histological and histochemical effects of tribendimidin on N. americanus have also been studied by the same investigators [86].
The in vitro response of adult (males and females) and free-living stages of N. americanus to a wide spectrum of anthelminthics was assessed by Kumar [87]. The observed activity varied from 0.0002 and 0.0007 mg/l for pyrantel pamoate and tricofenol piperazine, respectively, to about 8.47 and 7.6 mg/l for morantel tartrate and amoscanate, respectively. An overall conclusion was drawn that neither sex nor life cycle stage alone may serve as effective criteria for screening of anthelminthic agents. Furthermore, female parasites should be taken into account for the assessment of ED 50 values since they had required relatively the highest ED 50 levels in nearly all of the compounds studied. . Increasing the dose of mebendazole to 1.0 g also seemed not to improve the cure rate against hookworm infections [64]. These results are in contrast to those of Cauwenberg [88] that showed that increasing the single dose of mebendazole from 200 mg to 500 mg and to 600 mg yielded cure rates of 8%, 54% and 78%, respectively. One possible explanation for the reported discrepancies may lie in the different way cure rates have been measured.

Differential efficacies of mebendazole and albendazole
Vandepitte [89] has presented clinical evidence demonstrating that when a lower total dose of 600 mg mebendazole is administered as a three-day course for the treatment of hookworm infections, the cure rate exceeded 90%. This may indicate that the duration of exposure of the parasite to mebendazole rather than the concentration of the drug in the bowel is important for efficacy. However, as shown by Holzer and Frey [64], this was not the case for albendazole.
The reason for the difference in the clinical efficacy of mebendazole and albendazole is not very clear [64]. It may be due to the different pharmacokinetic properties of the two drugs. For example, while albendazole is well absorbed from the gastrointestinal tract and has systemic effect [90], mebendazole is poorly absorbed and has mainly intraluminal activity [91].

Drug resistance of N. americanus
Preliminary results from a placebo-controlled randomised trial conducted in Mali have shown that a single dose of 500 mg mebendazole was ineffective in the treatment of 103 patients infected with N. americanus [92]. Thus, there has been less than significant reduction of the parasite burden as assessed by faecal egg counts (18.5% eggs per gram of faeces). This failure of mebendazole to treat necatoriasis in the southern region of Mali is indicative that, among the other possibilities, the emergence of resistance to mebendazole by N. americanus may have contributed to drug failure [92]. By comparison, treatment with pyrantel pamoate (10 mg/kg) led to a marked reduction (75%) of faecal egg counts. Among other possibilities, the results may suggest the development of resistance to benzimidazole agents by N. americanus. It should be noted that benzimidazole resistance is widespread among nematode parasites in domestic animals. There have also been pronounced regional differences with extensive drug resistance being reported in South Africa, New Zealand and Australia [93,94]. Among the possible reasons for the development of resistance towards benzimidazoles, too frequent treatment, the use of inappropriate doses and the failure to provide alternative treatment strategies with other drugs are believed to be responsible for the resistance to specific anthelminthics [95,96]. It has been shown that resistance to benzimidazoles due to a loss of the drug's high affinity for binding to tubulin [52] had developed in intensively drug-treated livestock.
While little is known about selection pressure for appearance of drug resistance among human nematodes, studies on ovide parasites have shown that only a few rounds of treatment, at a time when most parasites are in their hosts, selected strongly for resistance [97]. Even though drug resistance may be slow to appear [98], it has long been anticipated because of the widespread misuse of anthelminthics in the therapy of human infections [97,99]. It is also possible that N. americanus in different localities may differ in its sensitivity to anthelminthics (e.g., mebendazole) due to genetic drift among the parasites relative to other geographically distant hookworm populations [92].
Among the other hypotheses for drug resistance that were proposed, different drug formulations, such as size particles, may have been responsible for the observed lack of activity [100]. In addition, Weshe et al. (101) have attributed the contrasting efficacy of two formulations of mebendazole to the efficiency of tablet breakdown in aqueous medium.

Hookworm infections and iron-deficiency anaemia
Hookworm infections in man, especially in children, are one of the leading causes of iron-deficiency anaemia [1,13,102-104] resulting directly from intestinal capillary blood loss following the feeding activities of fourth-stage larva and adult worms [105]. During the attachment of adult hookworms to the gastrointestinal mucosa, intestinal capillaries of the lamina propria are lacerated and the extravasated blood is either ingested by the fastened worm or will leak at the site of the parasite attachment [106]. The use of radioactive tracers [107] has allowed investigators to estimate that 30 µl of blood daily is lost to an individual N. americanus and 260 µl to A. duodenale [108,109]. Tissue damage and collagen exposure would normally trigger haemostasis. However, hookworms have evolved strategies to prevent haemostasis [110,111], such as the ability of N. americanus to inhibit Stypven clotting time and the coagulation Factor Xa (a pivotal component of the clotting cascade) [112][113][114][115][116]. Another important factor contributing to parasite survival is the ability of adult hookworms to interfere with platelet aggregation and activation, as well as to degrade fibrinogen [12,114,[117][118][119][120][121][122][123][124][125][126][127][128][129]. These complementary antihaemostatic strategies coupled with the accumulation of worms throughout the lifetime of the host [130] have enabled the parasites to establish a chronic infection commonly observed in human populations. Finally, the cumulative blood loss will usually lead to iron-deficiency anaemia in the infected individual [130].
Anaemia resulting from iron deficiency is a slowly developing condition that is known to proceed through progressive states of pre-latent, latent and manifest iron deficiency [131,132]. The pre-latent iron deficiency involves the loss of sequestered iron reserves (storage iron depletion) without a decrease in the iron supply to the developing red cells. The second latent stage (also referred to as 'iron-deficient erythropoiesis') is a corollary of diminished erythroid iron supply and occurs in the absence of a significant effect on circulating haemoglobin levels. The continued blood loss and lack of iron replacement then lead to the development of the third stage, manifested as iron-deficiency anaemia. All three stages can be monitored respectively by measuring the levels of serum ferritin, free erythrocyte protoporphyrin and haemoglobin [131,132].
In a number of studies, the relationship between hookworm infection and anaemia has been established by using hookworm egg production as an indirect method to measure the worm burden. Such studies have shown the degree of anaemia to be related to hookworm egg production, at least when egg counts exceeded 5000 eggs/g faeces [108].
Pritchard et al. [133] assessed the relationship between the iron status and the intensity of infection by monitoring the effect of N. americanus burden on the haemoglobin, haematocrit and the serum ferritin levels. The results have demonstrated the presence of a significant negative correlation between the plasma ferritin level and the hookworm burden, which was most pronounced in male patients. In contrast, there has been no correlation between plasma ferritin and hookworm egg count and no consistent correlation between the haemoglobin level or haematocrit and either measure of hookworm intensity. The overall results have indicated that the role of hookworm (N. americanus) in the aetiology of anaemia may be difficult to assess without the accurate measurement of the hookworm burden.
Stoltzfus et al. [134] have measured the intensity of hookworm infection using faecal egg counts, whereas gastrointestinal blood loss was measured by determining the concentration of haem in the faeces utilising the Hemoquant method [135]. The latter is used to determine the amounts of porphyrins in the faeces before and after digestion that degrades haem to its porphyrin constituents. Next, the concentration of porphyrin after digestion is used to calculate the total faecal haem concentration. The ratio of the porphyrin concentration before and after digestion will indicate where the blood loss occurred in the intestinal tract. A high proportion of porphyrin in the faeces before the digestion process will indicate blood loss higher in the intestinal tract [12].
It is estimated that a concentration of 10 mg/g faecal haem represents a daily loss of > 2.0 mg of iron, which is more than double the median iron requirement of a healthy school-age child [12]. In a disturbing trend, in children whose faecal haem exceeded 10 mg/g faeces, 100% had serum ferritin levels < 18 µg/dl, 93% had haemoglobin concentrations of < 110 g/l and 29% had haemoglobin concentrations < 70 g/l [134]. In this particular sample of children, the predominant hookworm was N. americanus, with about 10% having A. duodenale. However, when compared, blood loss from A. duodenale has been 2 -10 times higher than that caused by N. americanus [108].
Genta and Woods [136] have described a case of a patient with severe iron-deficiency anaemia in whom adult hookworm females were visualised, recovered and speciated during an upper endoscopic procedure.
By drawing attention to the importance of the regulation of stored iron levels in the process, Crompton and Whitehead [137] proposed a mathematical model to investigate how hookworm infection might disturb human iron metabolism and to explain how the iron metabolism may respond and adapt to a hookworm infection of varying degree.

Cutaneous larva migrans
Cutaneous larva migrans is a well recognised, usually self-limiting condition caused by the infectious larvae of nematodes. CLM is characterised by skin eruption and represents a clinical description rather than a definitive diagnosis [138]. Of the hookworm nematodes, the dog or cat worm, usually A. braziliense and A. caninum are the most common parasites causing CLM, although other species have also been implicated [139]. Even though CLM is diagnosed simply, its causative organism is not identified since biopsy is not thought to be beneficial [140].
Although A. braziliense is the hookworm nematode more responsible for CLM, N. americanus has been implicated in several cases of CLM [138,141,142] as well. The usual cutaneous manifestation of infection caused by N. americanus is pruritic reaction at the site of inoculation known as 'ground itch', followed within hours by an erythematous papular or papulovesicular rash [139]. The cutaneous symptoms will usually subside as the larvae penetrate the skin and migrate through the venous system into the lungs [138].
Topical treatment with thiabendazole suspension has been widely used because of its efficacy and safety [143]. Thiabendazole is usually applied at 25 mg/kg twice daily (to a maximum of 3.0 g per day) for 2 -5 days in adults and children [3].
Loughrey et al. [138] have applied a three-day course with oral albendazole (400 mg daily) to successfully treat CLM caused by the filariform larvae of N. americanus; the symptoms were resolved completely within two weeks and the patient suffered no side effects [144,145]. However, recurrence after a three-day course has been reported to occur due to heavy infestation [146]. By comparison, oral thiabendazole although equally effective, has shown more side effects, including nausea, vomiting and dizziness [147]. Ivermectin was also used in the therapy of CLM with cure rate as high as 100% with no toxic side effects [148,149].

Expert opinion
The global occurrence of hookworm infections by some estimates [97,150] reaches 900 million people. These infections are responsible for the development of about 1.5 million cases of iron-induced anaemia (hookworm anaemias), representing roughly one-third of all iron-deficiency anaemias to be of this type. In addition, cases of hookworm anaemias result in 30,000 -60,000 deaths each year. These figures taken together emphasise the public health importance of hookworm infections as a cause of serious morbidity and mortality, especially in millions of children where they may retard growth and intellectual development; a health problem that has been largely ignored by healthcare providers and researchers.
The highest priority in hookworm control is the reduction of mortality and morbidity by a combination of community-based chemotherapy and vast improvement in sanitation to prevent parasite transmission. Since the 1970s, two benzimidazole anthelminthic agents, mebendazole and albendazole have had established roles in the chemotherapy of hookworm infections in general and necatoriasis in particular. Recently, however, there has been evidence for the emergence of resistance to these two drugs. The latter, coupled with the fact that gaining access to therapy is problematic in many regions of the world and because dormant larvae are able to generate new infections months after chemotherapy has ceased, should provide impetus for renewed efforts towards the discovery of new chemotherapeutic agents. The need for new antihookworm therapies has been made even more acute by the fact that humans afflicted with an initial hookworm infection apparently do not acquire the strong protective immunity against future infections.
Lately, there has also been a renewed interest supported by new findings suggesting excellent possibilities for the discovery and development of vaccines against hookworm infections [1,151]. Thus, the development of genetically engineered vaccines would greatly help in the control of these infections, especially in highly endemic areas. In recent studies, recombinant polypeptides belonging to the Ancylostoma secreted protein (ASP-1) family have shown promise for reducing hookworm burdens after larval challenge infections in mice [151]. Since any vaccine is antibody-dependent, it is anticipated that a cocktail of different recombinant hookworm antigens may be required in order to effectively prevent heavy hookworm infections and disease. However, efforts to develop antihookworm vaccine have been hindered by the lack of a suitable animal model of hookworm infection resembling the human disease, as well as the need for easily available native hookworm antigens. In addition, useful serological correlates of antihookworm immunity are still poorly defined [151].