Intestinal Protozoa

Epidemiology

E. histolytica was first linked causally to amebic colitis and liver abscess by Lösch in 1875, and it was named by Schaudinn in 1903 for its ability to destroy host tissues. In 1925, Emil Brumpt proposed the existence of a second, morphologically identical but nonpathogenic Entamoeba species, E. dispar , to explain why only a minority of people infected with what was then termed E. histolytica developed invasive disease. Although Brumpt’s hypothesis was not accepted during his lifetime, it is now clear that he was correct, and E. histolytica (Schaudinn, 1903) has been formally reclassified as 2 morphologically indistinguishable species: E. histolytica , the cause of invasive amebiasis, and E. dispar , a nonpathogenic intestinal commensal parasite (see later section).

E. histolytica is a parasite of global distribution, but most of the morbidity and mortality from amebiasis occurs in Central and South America, Africa, and the Indian subcontinent. Fortunately, most of the 500 million persons worldwide previously believed to be asymptomatic E. histolytica cyst passers are actually infected with E. dispar . The total burden of disease caused by E. histolytica is unknown, however, a multisite molecular epidemiologic study conducted in Africa and Asia indicated that it accounted for roughly 2% of life-threatening diarrhea in children 12 to 23 months of age, and that E. histolytica –associated diarrhea is associated with stunted growth and development. Interestingly, the incidence of amebiasis appears to be rapidly decreasing in some locations, possibly as a result of modest improvements in infrastructure and/or to widespread use of metronidazole and the absence of animal reservoirs of E. histolytica . For example, in Dhaka, Bangladesh, where as recently as 2006 the majority of children were infected with E. histolytica at least once by age 4 years, amebiasis now rarely occurs.

E. histolytica has a simple, 2-stage life cycle that consists of an infectious cyst and a motile trophozoite. The cyst form measures 5 to 20 μm in diameter and contains 4 or fewer nuclei. The ameboid trophozoite, which is responsible for tissue invasion, measures 10 to 60 μm and contains a single nucleus with a central karyosome ( Fig. 113.1 ). The cysts are relatively resistant to chlorination and desiccation, and they can survive in a moist environment for several weeks.

Infection occurs following ingestion of cysts in fecally contaminated food or water. Within the lumen of the small intestine, the quadrinucleate cyst undergoes nuclear and then cytoplasmic division, giving rise to 8 trophozoites. Only about 10% to 20% of infected persons develop invasive disease, characterized by invasion of the colonic epithelium by trophozoites. Trophozoites that gain access to the bloodstream can spread hematogenously to establish infection at distant sites (most commonly liver abscess, as discussed in Chapter 84 ). Why some persons develop invasive disease and others remain asymptomatic remains a mystery, but environmental, host, and parasite differences are all likely to be important.

A molecular epidemiologic study that used PCR to amplify a polymorphic region of the E. histolytica genome and assign a genotype to different clinical isolates has demonstrated a correlation between different E. histolytica strains and the outcome of infection. The specific underlying genetic differences among strains that are responsible for altered virulence, however, remain unknown. On the host side, susceptibility to both intestinal and hepatic amebiasis is linked to HLA class II alleles, and a mutation in the leptin receptor that alters epithelial cell signaling is associated with E. histolytica infection in children and in a mouse model of colitis.

Pathogenesis, Pathology, and Immunology

Both amebic factors and the host’s inflammatory response contribute to tissue destruction during invasive amebiasis. Microscopy studies have defined a stepwise progression of disease ( Fig. 113.2 ). After excystation within the lumen of the small intestine, trophozoites adhere to colonic mucins and epithelial cells, followed by degradation of colonic mucus by secreted cysteine proteinases. Amebic killing of colonic epithelial cells occurs via multiple mechanisms, including contact-dependent induction of apoptosis and necrosis, and a phenomenon of cellular “nibbling” termed trogocytosis. The resultant acute inflammatory response causes additional tissue damage. Following epithelial destruction, extracellular matrix degradation and tissue invasion occur at sites of amebic invadopodia, which are cellular structures similar to those used by cancer cells during tissue invasion and metastasis.

The cecum and ascending colon are affected most commonly, although in severe disease the entire colon may be involved. On gross examination, pathology can range from only mucosal thickening to multiple punctate ulcers with normal intervening tissue ( Fig. 113.3 ) to frank necrosis. For unknown reasons, the downward invasion of amebic trophozoites often is halted at the level of the muscularis mucosa. Subsequent lateral spread of amebae undermines the overlying epithelium, resulting in the clean-based, flask-shaped ulcers so characteristic of classic amebic colitis. Early in infection, an influx of neutrophils is typical, but in well-established ulcers, few inflammatory cells are seen. Organisms may be seen ingesting red blood cells (erythrophagocytosis) ( Fig. 113.4 ). At distant sites of infection (e.g., liver abscess), similar pathologic characteristics also include central liquefaction of tissue surrounded by a minimal mononuclear cell infiltrate.

More than 90% of persons colonized with E. histolytica spontaneously clear the infection within a year. Children with fecal anti-amebic lectin immunoglobulin (Ig)A have short-lived protection from subsequent intestinal infection, but the protective role of secretory IgA is not certain. Furthermore, antibodies alone are unable to clear established infection, because asymptomatic cyst passers remain infected for months after anti-amebic antibodies develop. Reports that patients who are receiving glucocorticoids may be at increased risk for severe amebic colitis suggest that cellular immunity also plays an important role in control of E. histolytica infection. Despite this concern, no increase in disease severity has been observed in patients with AIDS. In fact, in a mouse model of amebic colitis, disease was exacerbated by CD4 ⁺ T cells.

Clinical Features

Infection with E. histolytica results in one of 3 outcomes. Approximately 80% to 90% of infected persons remain asymptomatic. The other 10% to 20% of infections result in invasive amebiasis characterized by dysentery (amebic colitis) or, in a minority of cases, extraintestinal disease (most commonly amebic liver abscess [see Chapter 84 ]).

In the USA, immigrants from or travelers to endemic regions, male homosexuals, and institutionalized persons are at greatest risk for amebiasis. In addition, malnourished patients, infants, older adults, pregnant women, and patients receiving glucocorticoids may be at increased risk for fulminant disease. When one or more of these epidemiologic risk factors are present, amebic dysentery should be considered in the differential diagnosis of occult or grossly bloody diarrhea.

The major diagnostic challenge for the clinician seeing a patient with amebic colitis is to distinguish the illness from other causes of bloody diarrhea. The differential diagnosis includes causes of bacterial dysentery, such as Shigella , Salmonella , Campylobacter species, and enteroinvasive or enterohemorrhagic Escherichia coli , and noninfectious diseases, including IBD and ischemic colitis. In contrast to bacterial dysentery, which typically begins abruptly, amebic colitis begins gradually over one to several weeks ( Table 113.1 ). Although more than 90% of patients with amebic colitis present with diarrhea, abdominal pain can occur without diarrhea; abdominal pain, tenesmus, and fever are highly variable. Weight loss is common because of the chronicity of the illness. Microscopic blood is present in the stool of most patients with amebic dysentery.

The most feared complication of amebic dysentery, acute necrotizing colitis with toxic megacolon, occurs in 0.5% of cases. This complication manifests as an acute dilatation of the colon, and 40% of patients die from sepsis unless it is promptly recognized and treated surgically. Unusual complications include the formation of enterocutaneous, rectovaginal, and enterovesicular fistulas and ameboma (a tumorous mass) due to intraluminal granulation tissue that can cause bowel obstruction and mimic carcinoma of the colon.

Dysentery resolves prior to presentation in most patients with amebic liver abscesses, although a history of dysentery is common in these patients. Extraintestinal sites of infection typically result either from direct extension of liver abscesses (e.g., amebic pericarditis or lung abscess) or from hematogenous spread of disease (e.g., brain abscess).

Diagnosis

Because amebiasis patients erroneously diagnosed with and treated for IBD with glucocorticoids can develop fulminant colitis, accurate initial diagnosis is critical. The gold standard for diagnosis of amebic colitis remains colonoscopy with biopsy, and colonoscopy should be performed whenever infectious causes of bloody diarrhea are strong considerations in the differential diagnosis of UC. Because the cecum and ascending colon are affected most often, colonoscopy is preferred to sigmoidoscopy. Classically, multiple punctate ulcers measuring 2 to 10 mm are seen with essentially normal intervening tissue (see Fig. 113.3 ); however, the colonic epithelium might simply appear indurated with no visible ulcerations; appear like UC with a myriad of ulcerations and granular, friable mucosa, or as a “poached egg” with a solitary mucus-covered ulcer; in severe cases where the ulcers have coalesced, the epithelium may appear necrotic. Histologic examination of a biopsy specimen taken from the edge of an ulcer reveals amebic trophozoites and a variable inflammatory infiltrate (see Fig. 113.4 ). Identification of amebae can be aided by periodic acid-Schiff staining of biopsy tissue, which stains trophozoites magenta.

Stool examination for ova and parasites, the traditional method for diagnosing amebiasis, should not be relied upon. Although the presence of amebic trophozoites with ingested erythrocytes strongly correlates with E. histolytica infection, these rarely are present, and in the absence of hematophagous trophozoites, microscopy cannot distinguish E. histolytica from E. dispar. Difficulty in distinguishing E. histolytica from nonpathogenic amebae (see later) and WBCs also limits the specificity of stool microscopy. The sensitivity of microscopy for identification of amebae is at best 60%, and may be reduced further by delays in the processing of stool samples. The primary utility of stool microscopy for ova and parasites in a patient with diarrhea, therefore, is to evaluate the stool for other parasitic causes of diarrhea.

Noninvasive methods to accurately differentiate E. histolytica from E. dispar include stool culture with isoenzyme analysis, serum amebic-antibody titers, PCR, and an enzyme-linked immunosorbent assay (ELISA) that detects the amebic lectin antigen in stool samples. A multiplexed PCR kit that simultaneously detects 22 GI pathogens, including the parasites Giardia. intestinalis , Cyclospora cayetanensis , Cryptosporidium species, and E. histolytica, with a specificity of 97.1% for all targets, is approved for use in the USA and is now widely available. Fecal antigen detection ELISA tests that accurately distinguish E. histolytica from E. dispar offer a less technically demanding alternative that can be used readily in the developing world. Fecal antigen detection, when compared with the gold standard of stool culture followed by isoenzyme analysis, is more than 90% specific and more than 85% sensitive to diagnose intestinal amebiasis when fresh fecal samples are analyzed without delay. In other studies, the sensitivity of fecal antigen detection has been less impressive, emphasizing the need for rapid processing of stool samples. It also may be possible to use this antigen detection test to diagnose amebic liver abscess, because before treatment is initiated, amebic antigen can be detected in the serum of more than 90% of patients who have amebic liver abscess.

Because serum anti-amebic antibodies do not develop in patients infected with E. dispar , serologic tests for amebiasis accurately distinguish E. histolytica and E. dispar infection. From 75% to 85% of patients with acute amebic colitis have detectable anti-amebic antibodies on presentation, and convalescent titers develop in more than 90% of patients. For amebic liver abscess, 70% to 80% of patients have detectable antibody titers on presentation, and convalescent titers develop in more than 90% of patients. Because anti-amebic antibodies can persist for years, however, a positive result must be interpreted with caution. For persons with known epidemiologic risks (e.g., emigration from or prior travel to an endemic region), a positive result might simply represent infection in the distant past. In the setting of recent travel to an endemic region and a positive antibody titer, diagnosis is supported by an appropriate symptomatic response to anti-amebic treatment.

Treatment

Drugs for treatment of amebiasis are categorized as luminal or tissue amebicides on the basis of the location of their anti-amebic activity ( Table 113.2 ).

The luminal amebicides include iodoquinol, diloxanide furoate, and paromomycin. Of these, paromomycin, a nonabsorbable aminoglycoside, is preferred because of its safety, short duration of required treatment, and superior efficacy; its major side effect is diarrhea. Approximately 85% of asymptomatic patients are cured with one course of paromomycin, and, because it is nonabsorbable and has moderate activity against trophozoites that have invaded the colonic mucosa, it also might be useful for single-drug treatment of mild invasive disease during pregnancy.

The tissue amebicides include metronidazole, tinidazole, nitazoxanide, erythromycin, and chloroquine. Of these, metronidazole and tinidazole are the drugs of choice, with cure rates over 90%. Nitazoxanide also appears to be efficacious, with similar cure rates in several randomized placebo-controlled trials. Erythromycin has no activity against amebic liver disease, and chloroquine has no activity against amebic intestinal disease.

Because approximately 10% of asymptomatic cyst passers develop invasive amebiasis, E. histolytica carriers should be treated. For such noninvasive disease, treatment with a luminal agent alone is adequate (e.g., paromomycin 25 to 35 mg/kg/day in 3 divided doses for 7 days). Patients with amebic colitis should first be treated with an oral nitroimidazole (either metronidazole [500 to 750 mg 3 times daily for 10 days] or tinidazole [2 g once daily for 3 to 5 days]) to eliminate invasive trophozoites. Metronidazole and tinidazole are believed to be less effective against organisms in the colonic lumen, and subsequent treatment with a luminal agent such as paromomycin is recommended to prevent recurrent disease. It is also for this reason that the familiar tissue amebicides (e.g., metronidazole) are not recommended as first-line agents for treatment of asymptomatic infection. At the recommended doses of metronidazole and tinidazole, GI side effects including nausea and vomiting develop in approximately 30% of patients. Because of severe GI side effects, simultaneous treatment with a nitroimidazole and a luminal agent generally is not recommended.

Most patients with amebic colitis respond promptly with resolution of diarrhea in 2 to 5 days.

Despite conflicting reports on the safety of the nitroimidazoles for the developing fetus during pregnancy, women with severe disease during pregnancy should probably be treated without delay. As discussed in Chapter 84 , metronidazole (750 mg 3 times a day for 10 days) followed by a luminal agent is also the treatment of choice for amebic liver abscess.

Control and Prevention

Prevention and control of E. histolytica infection depends on interruption of fecal-oral transmission. Water can be made safe for drinking and food preparation by boiling it for one minute, by halogenation (with chlorine or iodine), or by filtration. In the USA and Europe, modern water treatment facilities effectively remove E. histolytica . The importance of safe drinking water is highlighted by an outbreak of amebiasis in Tbilisi, Republic of Georgia, where there was a water-borne epidemic due to decay of the water treatment facilities following the demise of the Soviet Union. In the vast majority of the developing world, however, no modern water treatment facilities exist, and none are likely to be constructed in the foreseeable future. Naturally acquired immunity to intestinal amebiasis provides short-lived protection against reinfection, giving hope that a vaccine may be feasible. Because humans and some higher nonhuman primates are the only known hosts for E. histolytica , a vaccine that successfully prevents colonization might enable eradication of the disease.

Epidemiology

G. intestinalis (also called G. lamblia and G. duodenalis ) is a ubiquitous flagellated intestinal protozoan. Van Leeuwenhoek accurately described its motile trophozoite form in his own stools in 1681, but it was not until 1915 that Stiles named the species.

The life cycle of Giardia consists of an infectious cyst form and a motile trophozoite ( Fig. 113.5 ). The cyst is oval (8 to 12 μm long by 7 to 10 μm wide), contains 4 nuclei, and has a rigid outer wall that protects it from dehydration, extremes of temperature, and chlorination. Giardia cysts can survive in cold water for several weeks. Ingestion of as few as 10 to 25 cysts can result in infection. After ingestion, excystation occurs following exposure to stomach acid and intestinal proteases, each cyst giving rise to 2 trophozoites. Giardia trophozoites are pear-shaped (10 to 20 μm long by 7 to 10 μm wide), contain 2 nuclei, have 8 flagella for locomotion, and replicate by binary fission. The trophozoites live in the duodenum, where they adhere to enterocytes. Eventually they encyst, following exposure to alkaline conditions or bile salts, and are excreted in the stool to complete their life cycle.

G. intestinalis , which was defined originally as a species by morphology, is more accurately defined as a species complex with at least 8 major genotypes (assemblages A through H). Of these, only assemblages A and B have the broadest host range and are the only assemblages definitively known to infect humans. Both of these genotypes also commonly infect cats and dogs, highlighting the importance of these pets as reservoirs for human disease. Assemblage A isolates may be more virulent than assemblage B isolates.

G. intestinalis is the most commonly identified intestinal parasite in the USA and was identified in 7.2% of stool samples examined by state health departments in 1987. Giardiasis occurs in both endemic and epidemic forms via water-borne, food-borne, and person-to-person transmission. Worldwide, Giardia infects infants more commonly than adults, and in highly endemic regions, essentially all children are infected by 2 to 3 years of age. Giardia infection is associated with chronic diarrhea in children, but evidence that it causes acute diarrhea is controversial; some reports even suggest that it protects against acute diarrhea. Nevertheless, even asymptomatic Giardia infection is associated with malnutrition and impaired growth. In the USA, children in daycare and sexually active homosexual men have the greatest risk of infection. During a year-long longitudinal study at a USA daycare center, Giardia cysts were identified at some time in the stool of more than 30% of children. Additional risk factors for infection include drinking untreated surface water, a shallow well as a residential water source, swimming in any natural body of fresh water, and contact with a person who has giardiasis or contact with a child in daycare.

Pathogenesis, Pathology, and Immunology

Giardia causes malabsorptive diarrhea by an unknown mechanism. Trophozoites adhere (perhaps by suction) to the epithelium of the upper small intestine using a disk structure located on their anterior ventral surface. There is no evidence that trophozoites invade the mucosa, but electron microscopy has shown they damage the mucosal brush border. On biopsy, pathologic changes range from an entirely normal-appearing duodenal mucosa (except for adherent trophozoites), as was found in more than 96% of biopsy specimens in one large study, to severe villus atrophy with a mononuclear cell infiltrate that resembles celiac sprue. The severity of diarrhea appears to correlate with the severity of the pathologic change.

The host immune response plays a critical role in limiting the severity of giardiasis. When infected with Giardia , persons with common variable immunodeficiency develop severe, protracted diarrhea and malabsorption with sprue-like pathologic changes that resolve with treatment. Both systemic and mucosal humoral immune responses can be measured consistently following Giardia infection. High titers of anti- Giardia IgM, IgG, and IgA can be detected in the serum, and anti- Giardia secretory IgA can be detected in the saliva and in breast milk of infected mothers. Animal studies suggest that both early and late immune responses are important for control of Giardia infections. IL-6 is important in the early immune response to Giardia in mice, as are mast cells, which might function as IL-6 producers or via another mechanism. In a B–cell-deficient transgenic mouse model, infection with Giardia does not resolve, confirming the importance of the humoral immune response for clearance of established infections. In culture, Giardia trophozoites vary expression of a group of ∼200 cysteine-rich surface proteins termed variant surface proteins . In experimental human infections, G. intestinalis isolates have been shown to undergo antigenic variation, with switching of the predominant variant surface protein expressed after approximately 2 weeks, roughly the time required to mount an initial antibody response. Although the role of the variant surface proteins remains undetermined, antigenic variation might enable Giardia to evade the host immune response.

The importance of a cellular immune response also is clear from animal studies. Athymic nude mice are unable to control Giardia muris infection, but reconstitution with immune spleen cells results in partial control. Upon immune reconstitution, however, severe inflammatory changes and villus atrophy develop in the intestine, suggesting that the immune response to infection also might contribute to pathologic findings.

Clinical Features

Clinical manifestations of Giardia infection are highly variable, and range from an asymptomatic state to severe, chronic diarrhea with malabsorption. As noted earlier, Giardia infection in children is associated with chronic diarrhea, but epidemiologic evidence that it causes acute diarrhea in children in the developing world is limited. In one large study of biopsy-proved giardiasis, only 32% of patients had diarrhea; most had nonspecific GI complaints. Reported symptoms, in order of decreasing frequency, include diarrhea, fatigue, abdominal cramps, bloating, malodorous stool, flatulence, weight loss, fever, and vomiting ( Table 113.3 ). During a food-borne outbreak, the mean duration of diarrhea was 16 days, but symptoms resolved spontaneously in nearly half of infected patients after 7 to 8 days. Many patients with clinically apparent giardiasis suffer from lactose intolerance, malabsorption, or both for months following cure of infection.

As mentioned earlier, the severity of illness depends upon host and parasite factors. Different Giardia isolates have dramatically different abilities to cause disease during experimental human infections, and a larger proportion of children infected with Giardia assemblage A than assemblage B present with diarrhea. Furthermore, certain populations, including children younger than 2 years and patients with hypogammaglobulinemia, are more likely to develop serious disease. Despite the importance of cellular immunity for controlling infection in animal models and the increased risk of Giardia infection among sexually active homosexual men, giardiasis is not more common, severe, or resistant to treatment in patients with AIDS, except perhaps when AIDS is advanced.

Diagnosis

Examination of concentrated, iodine-stained, wet stool preparations and modified-trichrome-stained permanent smears has been the conventional approach to identifying Giardia infections (see Fig. 113.5 B ). Because cysts and trophozoites are present only intermittently in the stool, however, the sensitivity of such testing is only about 50%, even with examination of multiple specimens. With direct sampling of duodenal contents, such as duodenal aspiration or the string test, sensitivity can be improved to approximately 80%. On small intestinal biopsy specimens, identification of trophozoites requires careful examination of multiple microscope fields to ensure accuracy (see Fig. 113.5 A ).

Numerous molecular tests based on ELISAs or direct immunofluorescent antibody microscopy now are widely available commercially to diagnose giardiasis in stool samples. These assay kits all work well and have sensitivities >90% and specificities approaching 100%. As noted earlier (under E. histolytica ), a sensitive and specific multiplex PCR test is now available. The available molecular tests are all preferable to traditional microscopy and duodenal sampling as initial tests to evaluate for Giardia infection. The primary role of endoscopy is evaluation for other pathologic conditions.

Treatment

Metronidazole (250 mg orally 3 times a day for 5 days) is the preferred treatment for giardiasis. At this relatively low dosage, metronidazole generally is well tolerated and is 80% to 95% effective at eradicating Giardia. The most common side effects of metronidazole are nausea, a metallic taste, and a disulfiram-like reaction upon consuming alcohol.

Nitazoxanide appears to be at least as effective as metronidazole and has the advantage of being available in a liquid formulation for use in pediatric patients. The recommended dosage in children is 100 mg (ages 12 to 47 months) or 200 mg (age >4 years) twice daily, and in adults is 500 mg twice daily for 3 days.

Alternative regimens include tinidazole (2 g orally for 1 dose), quinacrine (2 mg/kg 3 times a day for 5 days; maximum 300 mg/ day), furazolidone (100 mg orally 4 times a day for 7 to 10 days), or paromomycin (25 to 35 mg/kg/day in 3 divided doses for 7 days). Single-dose treatment with tinidazole has been used for years in Europe and the developing world and is approved by the FDA. Because paromomycin is not absorbed and there have been conflicting reports regarding the safety of metronidazole and tinidazole for the developing fetus, paromomycin may be especially useful for treatment of giardiasis during pregnancy.

As noted earlier, many patients have prolonged lactose intolerance following Giardia infection, which can mimic ongoing infection ; therefore, the diagnosis should be reconfirmed before repeating therapy. For people in whom treatment fails, repeat therapy with the same drug (e.g., with higher doses of metronidazole) or combination therapy with metronidazole and quinacrine might work. Nitazoxanide alone also may be effective. Patients in whom treatment repeatedly fails should be evaluated for common variable immunodeficiency.

Control and Prevention

Control of giardiasis relies on interruption of fecal-oral transmission. Water can be made safe for drinking and food preparation by boiling (for one minute), halogenation (with chlorine or iodine preparations), or filtration. Because of the low infectious dose of Giardia cysts and the poor hygiene of infants and children, person-to-person spread in daycare centers is much more difficult to control. Temporarily removing infected ill children from daycare is ineffective, perhaps because many infected children remain asymptomatic and infection goes unrecognized. In the developing world, endemic giardiasis is unlikely to be controlled until facilities become available for adequate filtration of water and disposal of sewage.

A Giardia vaccine composed of killed G. intestinalis trophozoites was licensed for use in cats and dogs, but, perhaps due to antigenic variation, was not very effective and is no longer manufactured. Recently, a recombinant vaccine prepared from a broad repertoire of isolated variant surface proteins was shown to be highly efficient in cats and dogs. Furthermore, immunization of dogs in a highly endemic area significantly reduced the prevalence of infected children, suggesting the possibility of blocking zoonotic transmission.