Epidemiology studies indicate an inverse relationship between endemic helminthiasis and autoimmune and allergic diseases

A number of authoritative and convincing review articles have documented increasing incidences of diseases such as inflammatory bowel disease (IBD), insulin-dependent diabetes mellitus, multiple sclerosis, celiac disease, asthma and atopic/allergic conditions in economically developed regions of the world (Elliott et al. 2000; Araujo et al. 2004a; Cooper, 2004; Hunter and McKay, 2004; Wilson and Maizels, 2004). These disorders are too diverse and their emergence too rapid to support an exclusively genetic explanation for their increased incidence. Thus, one is drawn to a consideration of changes in environmental or nutritional factors as risk factors in the occurrence of, for example, IBD. Indeed, the geographical distribution of the aforementioned diseases is remarkable as it virtually excludes areas where helminth infections are endemic. In fact, Weinstock and colleagues presented a very elegant case for the association of de-worming with increasing incidence of Crohn's disease, a major form of IBD (Weinstock et al. 2002). However, correlative studies must always be viewed with caution as the data can be confounded by study bias, differences in the criteria used to include patients in the study, under or misdiagnosis of the disease in question (Farrokhyar, Swarbrick and Irvine, 2001), and really cannot disentangle the issues of cause and effect: a strong significant relationship between one continuous variable and another does not mean that that the two events are causally linked. So while the epidemiological data are supportive of the postulate that infection with parasitic helminths could prevent or reduce the severity of specific diseases, evidence in support of a causative relationship must be sought and the mechanisms that account for any benefit derived from helminth infections defined.

In presenting the case for helminth infections providing a concomitant health benefit to their hosts, the deleterious effects of parasitic infections must not be overlooked. Helminth infections cause severe morbidity and mortality, are detrimental to physical and cognitive development and, among other concerns, are a leading cause of neurological disorders in developing nations (Oberhelman et al. 1998; Oyewole et al. 2002; Ito, Nakao and Wandra, 2003; Udani, 2005). Thus, 1470 million people are infected with the nematode whipworm Trichuris trichiura, and as a consequence suffer from dysentery, anemia, malnutrition and rectal prolapse (Compton, 1999; Saldiva et al. 1999).

The T helper (Th) cell paradigm

In the mid-1980s Mosmann and colleagues unequivocally showed that murine CD4⁺ T helper cell clones developed in vitro could be classified as Th1 or Th2 based on mitogen-evoked production of interferon-γ (IFNγ) and interleukin (IL)-4, respectively (Mosmann et al. 1986; Reiner, 2001). The Th1-Th2 paradigm is readily demonstrable in murine systems, and while a similar dichotomy likely exists in humans (Turner et al. 2003; Romagnani, 2004), a definite designation of Th1 or Th2 can be obscured depending on when the analysis is performed during the course of the infection (Allen and Maizels, 1997). Moreover, Th1 and Th2 cells are reciprocally cross-inhibitory; activation of a Th1 response (e.g. IFNγ production, activation of macrophages and CD8⁺ cytotoxic T cells, and generation of IgG_2α) would be accompanied by inhibition of the mobilization of Th2-dominated responses, such as IgG₁ and IgE production.

It is now clear that Th1 and Th2 cells are but two phenotypes among a spectrum of T cell types and that the generation and resolution of adaptive immune responses involves multiple types of T cell (McGuirk and Mills, 2002). Recognition of CD8⁺ T cells as a source of cytokines and identification of the regulatory B cell adds another layer of complexity to the adaptive arm of the immune system (Mosmann, Li and Sad, 1997; Velazquez, Wei and Braun, 2005). Thus any immune response, and particularly that against parasitic helminths, requires analysis from an integrated immunophysiological perspective, where the theme of connectedness between cells and systems is emphasized (Perdue and McKay, 1994). While definition of adaptive immunity as Th1- or Th2-type can be restrictive, the Th1-Th2 paradigm does nevertheless provide a convenient conceptual framework to begin to assess immune responses. In this context, assessment of serum, tissue or in vitro-stimulated immune cells from helminth-infected animals consistently reveals increases in Th2-type cytokines (i.e. IL-4, IL-5, IL-13 and IL-10) and Th2-driven immune effector systems (e.g. increased mucus-producing goblet cells in the intestine) (Grencis, 2003). However, with the exception of intestinal helminth infections, where the Th2-driven responses are clearly anti-helminthic, the relative value of Th2 responses to limiting immunopathopathology and tissue injury, helminth eradication, or contrarily helminth survival need to be viewed in a species-specific manner, with consideration given to the various life-cycle stages of the different parasites (Allen and Maizels, 1996). For instance, the cytokine response in mice infected with the parenteral cestode, Taenia crassiceps (a widely accepted model of human cycticercosis caused by Taenia solium), is of a Th1 or mixed Th1/Th2 following infection, and converts to Th2-dominated responses in older infections characterized by the presence of granuloma-encased cysticerci (Toenjes and Kuhn, 2003). Furthermore, serum measurement of cytokine levels in helminth-infected humans does not universally reveal an overtly Th2-skewed pattern. This may be a reflection of the time of cytokine determination after infection (i.e. acute verses chronic), or indicative of mixed parasite infections (and/or bacterial infections), and, of course, tissue levels of cytokines may not be the same as serum levels of cytokines. However, recent data from a study involving Cameroonians suggests that assessment of Th2 responsiveness from stimulated peripheral blood mononuclear cells can predict susceptibility to gastrointestinal nematodes (Jackson et al. 2004).

Parasitic helminths and the modulation of disease

Cytokine profiles in patients with Crohn's disease have led to the wide-spread acceptance of this disorder as a Th1-dominated condition (Sawa et al. 2003) Thus, helminth infections by mobilizing Th2-cyokine responses would be predicted to be of value in blocking Crohn's disease. As proof-of-principal in support of this postulate, mice infected with nematode (Heligmosomoides polygyrus, Trichinella spiralis), cestode (Hymenolepis diminuta) or trematode (Schistosoma mansoni) parasites, or treated with S. mansoni egg antigen, are protected from colitis induced by direct intra-rectal administration of tri- or dinitrobenzene sulphonic acid (TNBS, DNBS) or by addition of dextran sodium-sulphate (DSS) to their drinking water: these models of colitis are associated with increased Th1 cytokine levels (Reardon et al. 2001; Khan et al. 2002; Elliott et al. 2003; Moreels et al. 2004; Hunter et al. 2005). Similar experimental data can be found in support of parasitic helminths blocking other Th1-type diseases in mice, such as Helicobacter-induced gastritis and experimental autoimmune encephalomyelitis (a model of multiple sclerosis) (Fox et al. 2000; La Flamme, Ruddenklau and Backstrom, 2003). So, are these studies relevant to human disease? In a recent small non-blinded study, Summers and coworkers reported that consumption of ova from the porcine whipworm, Trichuris suis, led to disease remission, or significant symptom relief, for patients with Crohn's disease (Summers et al. 2005a). While preliminary, these data are intriguing and generally fit the Th1-Th2 paradigm.

These studies generate 2 important areas for discussion. The beneficial effects of helminth parasites in murine models of inflammation or Crohn's disease have, with the exception of H. polygyrus, been observed in non-permissive systems, suggesting that the immune response mobilized to evict the worm is of sufficient magnitude to block the initiation of colitis or reduce established disease. Indeed, unlike infection of the non-permissive murine host, rats develop a patent H. diminuta infection and are not protected from DNBS-induced colitis (Hunter et al. 2005). In the same study, mice deficient in the transcription factor, signal transducer and activator of transcription (STAT)-6, which do not expel H. diminuta (McKay and Khan, 2003), were not protected from DNBS-induced colitis by prior infection with H. diminuta. However, the possibility that STAT-6 is important in the anti-colitic effect has not been unequivocally excluded and indeed exogenous IL-4 (STAT6 is an important intracellular mediator of IL-4 bioactivity) has been shown to block TNBS-induced colitis (Hogaboam et al. 1997). To extrapolate these data to human disease, where helminth infections are often chronic, one can hypothesize that the immune response against the worm, while incapable of eliminating the parasitic burden is able to reduce Th1-mediated disease and pathology (I will return to the issue of humans with chronic helminth infections later). In the context of the therapeutic benefit gained by the patients with Crohn's disease who ingested T. suis eggs, it should be noted that these patients are unlikely to have a history of helminth infections and that T. suis will not establish a chronic infection in humans.

The corollary of the hypothesis that parasitic helminths will block Th1-driven disease by promoting an antagonizing Th2 response is that the helminth infections would predispose to Th2-dominated disorders. This in fact appears not to be the case: enhanced IgE production is characteristic of individuals with helminth infections, but this is not associated with increased allergic or atopic-type disorders (Yazdanbakhsh, van den Biggelaar and Maizels, 2001). Indeed, and seemingly paradoxically, asthma and allergy (Th2- and IgE-dominated conditions) are significantly less prevalent in areas of endemic helminth infections (Bashir et al. 2002; Yazdanbakhsh, Kremsner and van Ree, 2002). Also, in an assessment of H. diminuta amelioration of DNBS-induced colitis it was found that infected mice simultaneously sensitized to ovalbumin were not hypersensitive to subsequent ovalbumin challenge as gauged by mast cell-mediated responsiveness compared to animals only sensitized by ovalbumin (Hunter et al. 2005). Equally provocative are recent data showing that patients with ulcerative colitis, a disorder associated with Th2 cytokines, obtained some symptomatic relief after ingestion of T. suis eggs (Summers et al. 2005b) (in an accompanying editorial the novelty of this approach and the small therapeutic benefit that was achieved were highlighted (Mayer, 2005)). How then does one rationalize Th2-promoting infections with reduced incidence or development of diseases driven by Th2-type cytokines?

Helminth-evoked interleukin-10 production

Since strict adherence to the Th1-Th2 paradigm or the hygiene hypothesis does not universally explain the down-regulation of many diseases or immunopathological states by helminth infections, other explanations have been sought. A concept presented by a variety of investigators from diverse disciplines is that of creation of an immuno-suppressed or immuno-regulatory environment as a consequence of helminth infections. A variety of model systems suggest that IL-10 is important in the immune-regulation that accompanies helminth infection and is important for host survival (Schopf et al. 2002). Thus, the protection against DNBS-induced colitis afforded by H. diminuta infection is lost if the mice are simultaneously treated with an IL-10 neutralizing antibody (Hunter et al. 2005). Similarly, the anti-colitic effect of S. mansoni eggs correlated with increased tissue levels of IL-10 mRNA (Elliott et al. 2003). Furthermore, the protection that nematode or S. mansoni infection provides against atopic, allergic and autoimmune conditions can be associated with increased IL-10 levels and ablated by blocking IL-10 activity (van den Biggelaar et al. 2000; Araujo et al. 2004b; Wohlleben et al. 2004). Ramaswamy and coworkers extended the focus on IL-10, and suggested that part of the immuno-modulatory effect of IL-10 following S. mansoni infection was due to the subsequent production of prostaglandin (PG) E₂, which is generally immuno-suppressive (Ramaswamy, Kumar and He, 2000). T helper 2 cells could be the source of the IL-10, but it is also likely that regulatory T cells (T_regs) in helminth-infected individuals are a significant source of IL-10 (O'Garra and Vieira, 2004). In the intestine, T_regs have been shown to limit inflammation and tissue injury (Groux et al. 1997). Moreover, recent findings suggest that inhibition of fatal anaphylaxis in mice infected with S. mansoni was due to IL-10 from B cells (Mangan et al. 2004); and, there is an increasing body of literature concerning the regulatory B cell and more recently how B cells might modulate the activity of T_regs (Wei et al. 2005). A considerable amount of data can be presented in favour of IL-10 as a major player in helminth-evoked beneficial health effects; however, this is certainly not the full story. Interleukin-10 deficient mice develop a spontaneous colitis the severity of which can be reduced by H. polygyrus infection (Elliott et al. 2004). In this instance, IL-10 cannot be mediating the nematode-evoked effect. Also, the promise of IL-10 therapy for IBD as predicted by murine models of colitis has not been fulfilled in clinical trails, leaving one with the unsatisfactory explanations that either the recombinant IL-10 is not being administered in the appropriate manner or that Crohn's disease is not amenable to IL-10 therapy (Li and He, 2004).

Regulatory T cells and transforming growth factor-β

As mentioned above, the regulatory T cell is an important source of IL-10 but other phenotypically distinct T cells exist that exert their effects by direct cell-cell contact or soluble factors, notably IL-10 and TGFβ (O'Garra and Vieira, 2004). It is likely that analysis of the host immune response to all classes of helminth parasite will reveal increases in T_regs during the infection (and/or rejection). Given the ability of these cells to control immune reactions, it is conceivable that T_regs mobilized in response to helminth infections could, in a bystander manner, limit other pathophysiological reactions. To date, a limited number of studies have been presented in support of this hypothesis. For instance, IL-10 producing CD4⁺CD25⁺ T_regs have been described in S. mansoni infected mice, FoxP3 (a T_reg-specific transcription factor) has been shown in S. mansoni or H. polygyrus-infected mice and, a regulatory T cell designated as type 1 (Tr1) has been identified in patients with chronic onchocerciasis (Satoguina et al. 2002; Elliott et al. 2004; Hesse et al. 2004; McKee and Pearce, 2004).

Increases in TGFβ have been associated with immunomodulation following infection with protozoan parasites (Omer, Kurtzhals and Riley, 2000), and while recent reports have documented TGFβ production following infection with helminths, the contribution of this cytokine to inhibition of concomitant disease has not been tested (Omer et al. 2003; Paalangara, McClure and McCullagh, 2003).

Given the intense current interest in T_regs and TGFβ in the control of mucosal immunity, it is reasonable to speculate that substantial evidence in support of T_regs and TGFβ involvement in the amelioration of disease by helminth infections will be forthcoming. However, as is the case with IL-10 (and PGE₂), findings relating to mobilization of T_regs and TGFβ in helminth infections and any concomitant health benefits for human hosts must be placed in the context of infections followed by spontaneous cure, chronic infections and/or polyparasitism (see below).

Antigen-presenting cells: dendritic cells and macrophages

It has already been noted that B lymphocytes can be responsible for some of the effects that accompany helminth infections, and it would be naïve to exclude consideration of other immune cells. For example, primary immune responses are in large measure dependent on dendritic cells (DC). The activity of the DC, and the subsequent polarization of the immune cell response is dependent on the nature of the antigen, the expression of co-signals (e.g. B71 (CD80) verses B72 (CD86)) and other molecules (e.g. CD40), and the nature of the cytokine milieu (Maldonado-Lopez and Moser, 2001). The ability of DC to cause the differentiation of Th2 cells is increased by exposing the DC to products from Acanthocheilonema vitaea or S. mansoni (Whelan et al. 2000; Thomas et al. 2003). In addition, using gene-microarray analysis it was recently shown that soluble egg antigen from S. mansoni affected the expression of >100 LPS-regulated genes in DCs (Kane et al. 2004). These findings led the authors to conclude that helminth-derived products could affect toll-like receptor (TLR: LPS binds TLR-4) signalling and this would certainly impact on the nature of innate immunity and subsequent adaptive immune responses. However, while highly probable, direct involvement of DC subtypes in the modulation of disease by helminth infections has not been shown and needs to be specifically addressed.

Macrophages are important scavengers of abnormal/damaged self-tissue and foreign antigen, and while the respiratory burst associated with the phagocytic activity of macrophages can damage host tissue there is increasing evidence of additional macrophage phenotypes that can influence T cell polarization and may be more immunologically ‘quiet’ (Mills et al. 2000). Helminth infections have been associated with an ‘alternatively activated macrophage’ (AAM) that develops under the influence of IL-4 and IL-13 (Gordon, 2003). These cells are characterized by increased expression of arginase-1 compared to inducible nitric oxide synthase (iNOS) in conventional macrophages, and by the expression of Fizz1 and Ym1 genes (Loke et al. 2000; Donnelly et al. 2005). Arginase-1 is active in tissue remodeling, since L-arginine metabolites are consumed during cell division and collagen deposition. Little is known of the activity of the Fizz1 or Ym1, but both have been implicated in tissue repair processes and Ym1 may be a chemoattractant for eosinophils (Welch et al. 2002; Liu et al. 2004; Nair et al. 2005). Thus, differentiation of AAMs could be a reaction to the tissue damage caused by migrating helminths or the abrasion/destruction associated with nematode feeding. Moreover, the AAMs that develop in response to Brugia malayi block lymphocyte proliferation by a cell-to-cell contact that is not due to the induction of apoptosis (Loke et al. 2000), while S. mansoni-induced T cell anergy was associated with up-regulation of programmed death ligand-1 on macrophages (Smith et al. 2004). Also, peritoneal macrophages that develop in chronically T. crassiceps-infected mice display increased PGE₂ production but dampened nitric oxide synthesis upon stimulation: these cells promoted the development of IL-4-producing CD4⁺ T cells (Rodriguez-Sosa et al. 2002). A similar non-classical macrophage phenotype has been identified in airways tissue following T. spiralis infection (Dzik et al. 2004). Thus, AAMs could contribute to the suppression of concomitant disease in helminth infected individuals; while this is a reasonable speculation, there is currently no data to support the supposition.

Major challenges in assessing helminth infections as a benefit to blocking other diseases in their hosts, will be to provide a precise evaluation of the relative individual contributions of the immune cells and the collective effect of mobilizing multiple immuno-regulatory cells (plus their respective mediators), and understanding how the cells/mediators interact to combat helminth infections and provide additional health advantages for the host.

The neuroimmunophysiological perspective

Pioneering work by Castro in the 1980s drew attention to the fact that the host response to a parasitic infection is an integrated one, involving a variety of effector cells and communication between cells that hitherto had not been well appreciated (Castro, 1989). In this context one must look beyond the ‘classical’ components of the immune system and ask, what other cells or mediators might be involved in the response to helminth infections?

Increased mucus production and water secretion are characteristic of intestinal helminth infections, particularly those in which spontaneous cure is evident (McKay et al. 1990b; Masson et al. 1996). For example, H. polygyrus-induced IL-4 in mice had direct and indirect (via nerves) effects on epithelial ion transport especially following secondary worm infection (Shea-Donohue et al. 2001). Enhanced mucus and water secretion would reduce the contact time between luminal agents (e.g. microbial or food antigens) and the gut epithelium, ‘wash’ the surface of the enterocyte and contribute to the flushing of noxious agents from the gut. As a consequence, aggravated intestinal disease evoked by luminal stimuli would be reduced. Similarly the altered muscle contractility and increased peristalsis that accompanies enteric helminth infections would result in enhanced caudally-directed movement and expulsion of antigen and parasites from the body (Vallance, Blennerhassett and Collins, 1997; Zhao et al. 2003).

Muscle contraction is under neuronal control and there is abundant information on bi-directional communication between nerve cells and all classes of immune cells. Some neurone-derived molecules enhance immune responses (e.g. substance P) (Suzuki et al. 1999) and others dampen immune reactivity (e.g. vasoactive intestinal peptide (VIP)) (Delgado and Ganea, 2001). In fact, descriptions of neuronal changes following helminth infections are common; however, the functional significance of the altered neuronal activity is often less clear (McKay and Fairweather, 1997). Nerve growth factor (NGF) affects not only neuronal growth but also a variety of immune functions (Bonini et al. 2003). Murine recombinant Fizz1 can inhibit NGF-mediated survival of rat dorsal root ganglion neurones in vitro (Holcomb et al. 2000). This AAM/Fizz1-Neuron/NGF relationship is intriguing and allows for the possibility that macrophage-nerve communication could influence neurogenic inflammation or the participation of nerves in mucosal immune responses. Similarly, T. spiralis infection has recently been shown to block up-regulation of iNOS expression via an IL-4/STAT6-dependent, T cell-independent mechanism (Bain et al. 2005): a predicted consequence of this would be reduced enteric inflammation. Juxtaposition of the facts that neuronal changes accompany helminth infections and that immune responses can be modified by neurochemicals/neuropeptides leads to the testable hypothesis that amelioration of inflammatory or autoimmune disease by helminth infections is due in part to changes in the local or systemic nervous system.

Similarly, enteroendocrine cells (EC) are a source of mediators that can affect mucosal immunity (e.g. serotonin, neurotensin). Increases in EC numbers can accompany helminth infection (McKay et al. 1990a) but the relevance of this to altered immune reactivity or the suppression of concomitant disease has not been determined. Enteroendocrine cells are potentially excellent surveillance cells: their luminal aspect is exposed to the gut lumen and they contain a variety of amines and neuropeptides that can be quickly released to act at both local (autocrine, paracrine) and distant (endocrine) sites. The role of enteroendocrine cells in enteric helminth infections deserves greater research efforts.

Active parasite involvement

To this point the beneficial effects of helminth infections have been considered as components of the hosts' response to the parasite, the inference being that this is solely an active host response. This is clearly not the case. At the outset it was noted that the successful parasite is one that survives the host's attempts to quickly identify and destroy/eradicate it. Behnke and colleagues presented some of the seminal observations on immuno-suppression caused by acute and chronic helminth infections (Behnke, Barnard and Wakelin, 1992) and there are innumerable examples of how specific species of helminths modulate host immunity and irrefutable evidence that helminths subvert host immune responses (Maizels et al. 2004). In the latter instance, active modulation of host immune cell activity could be a major contributor to any subsequent health benefits that coincide with helminth infections. For example, many helminth-derived excretory/secretory products have general immuno-suppressive properties (Arechavaleta, Molinari and Tato, 1998). Other helminth-derived molecules seem to preferentially influence the development of a restricted spectrum of immune events or cytokines – inhibition of IL-4 or potentiation of IFNγ, for example (Lawrence, Allen and Gray, 2000; Rigano et al. 2001). Finally, and of particular relevance in the light of immuno-regulation, protein components from Onchocerca volvulus, A. viteae, and H. diminuta and a lipid-based molecule from S. mansoni appear to specifically promote IL-10 production from host immune cells (van der Kleij et al. 2002; Hartmann and Lucius, 2003; Wang and McKay, 2005). So, unlike systems that assess the implications of delivering purified antigen (with or without adjuvant), the host's immune response, whether anti-helminthic, beneficial to the helminth, or advantageous to the host by limiting concomitant disease, must always be viewed from the perspective of host-parasite interactions.

Finally, there is increasing evidence that helminths can actively manipulate their hosts' immune response by the production of cytokine, or neuropeptide, analogues, including IFNγ, IL-12 (p40 chain), macrophage migration inhibitory factor (MIF), TGFβ, and vasoactive intestinal polypeptide (VIP) (Foster and Lee, 1995; Grencis and Entwistle, 1997; Gomez-Escobar, Lewis and Maizels, 1998; Zang et al. 2002; Wang and McKay, 2005).

Human helminthiasis

The host response to parasitic infection can often be presented from an immunologists' perspective, overlooking the parasitiological context. In areas with high prevalence of parasitic infections, individuals are often infected with multiple species of helminth and/or protozoal parasites. However, laboratory models (particularly rodents infected with gastrointestinal helminths or S. mansoni) that have been so instrumental in defining the stereotypic host immune responses to helminths almost exclusively use a single species of parasite. Indeed, it is known that concurrent infection of Nippostrongylus brasiliensis-infected rats with H. polgyrus prevents the spontaneous cure of N. brasiliensis, converting it to a chronic infection (Wescott and Colwell, 1980). This issue of polyparasitism is intriguing and has implications for worm fecundity (which is often higher in mixed helminth infections) and for immune responses in the host (Tchuem-Tchuente et al. 2003). Specific helminth co-infections have been assessed in mice (Behnke et al. 1992) but not in the context of how this would impact the initiation or progression of another disease. Similarly, while data is emerging on the health effects of polyparasitism in defined populations of humans (typically in developing regions of the globe), the full importance of this to suppression of other allergic or inflammatory diseases is not known.

The beneficial effect of parasitic helminths in alleviating colitis in some murine model systems was dependent on the spontaneous rejection of the helminth, but human helminth infections are typically of a chronic nature. In this context, there is evidence that chronic helminth infections do ameliorate disease. For instance, results of a randomized controlled clinical trial on the effect of eradiation of intestinal helminths (Ascaris and Trichuris) in Gabonese school children showed that the anthelminthic therapy lead to increased development of skin sensitivity to house dust mites (van der Biggelaar et al. 2004). In a related study, it was found that schistosome-infected children had reduced responsiveness to Toll-like receptor ligands (TLR) such as LPS, indicating altered signaling in the innate immune response (van der Kleij et al. 2004). This suggests that de-worming strategies could leave the individuals susceptible to not only atopic disorders but also to infection with microbes such as Plasmodium and Mycobacterium that are endemic in geographical regions of high helminth prevalence. However, de-worming could lead to reducing incidence of acquired immune deficiency syndrome (AIDS), since Bentwich and colleagues have suggested that helminth infections increase T cell numbers and their expression of the chemokine receptors that the AIDS virus uses to attach to and invade T cells (Kalinkovich et al. 1999; Borkow et al. 2000).

One must also question differences between a single ‘bolus’ of infection used in the laboratory model systems, with that of low-dose, trickle infections with the same species, or constant exposure to a variety of parasitic helminths – conditions representative of areas of endemic helminth infections. Indeed, trickle infection of rats with N. brasiliensis may result in a chronic infection and not the rapid rejection response typical of a high-dose infection (Jenkins and Phillipson, 1972). One would expect that this difference in infectivity would be reflected in either quantitatively or qualitatively different immune responses.

Such parasitiological issues are likely to be key to fully understanding how helminth infections modulate the induction and/or course of concomitant disorders in their hosts, and must be incorporated into models of helminth-alleviation of disease based on epidemiological observations. Similarly, these parasitiological issues should be borne in mind when considering the development of therapies based on data from helminth-infected rodents.

In making the case for the ‘beneficial helminth’, one would be remiss in not drawing attention to a number of caveats. As noted above, animal models are not entirely reflective of human-helminth interactions and so caution must be exercised in extrapolating data from the former to the latter. Also, the choice of helminth is important and obviously as innocuous a species as possible would be desired. It would make little sense to treat the debilitating IBDs with potentially fatal schistosomasis, although infection with non-permissive nematode or cestode species could be effective therapies. If actual helminth infection is considered as a therapy then the issue of palatability for the patient becomes pertinent. Additionally, one would advocate precise evaluation of the patient and their responsiveness to conventional therapies before considering the use of helminth infections. Importantly, one must be aware of the potential problems of introducing a parasite into an abnormal host as underscored by West's comments on the use of T. suis to treat patients with IBD (2005). In the abnormal host, the enteric parasite may be faced with a subtly or dramatically altered physio-chemical environment that might permit localization of adults, larvae or eggs to non-intestinal sites (e.g. the nervous system) and the development of pathology. It could take months or years for such adverse effects to appear and their incidence might be low, and so long-term follow-up of patients ‘treated’ with helminth parasites should be practiced.

So while actual prescription of helminth therapy may be feasible in the future, one can assume that it will come with some degree of concern and skepticism by the medical community; a viewpoint that can be swayed if helminth therapy is highly effective. However, the concerns outlined above relating to treating patients with ‘therapeutic helminths’ are less relevant in laboratory model systems where the goal is to understand the mechanism(s) by which helminth infections ameliorate diseases, and is so doing allow for the development of pharmacological agents that imitate the beneficial effect(s) of infection with parasitic helminths.