NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Baron S, editor. Medical Microbiology. 4th edition. Galveston (TX): University of Texas Medical Branch at Galveston; 1996.

Cover of Medical Microbiology

Medical Microbiology. 4th edition.

Show details

Chapter 79Intestinal Protozoa: Amebas

.

General Concepts

Entamoeba Histolytica

Clinical Manifestations

Patients have acute or chronic diarrhea, which may progress to dysentery. Extraintestinal disease may be present as a complication or as a primary problem (e.g., liver, lung or brain abscess, or skin or perianal infection).

Structure

The trophozoite is 10 to 60 µm in diameter, ameboid, actively motile, and often erythrophagocytic. In stained specimens, the nucleus has a central karyosome with finely beaded peripheral chromatin. The cyst form is rounded, 10 to 20 µm in diameter, with one to four nuclei showing the characteristic appearance. A chromatoidal bar with rounded or square ends may be seen.

Classification and Antigenic Types

Pathogenic strains can be grown at 37° C but not at room temperature and fall into specific enzyme assay groups.

Multiplication in the host occurs by binary fission. Nuclear replication produces four nuclei during cyst maturation. During excystation the cyst divides to form four cells which immediately divide again to yield eight tiny amebae.

Pathogenesis

The colon may be colonized without invasion of mucosa. The critical factor determining colonization is the ability of the ameba to adhere to colonic mucosal lining cells. Invasion of the mucosa produces ulcers that sometimes progress by direct extension or by metastasis. Metastatic infection first involves the liver. Extension or metastasis from the liver may involve the lung, brain, or other viscera.

Host Defenses

Gastric acid and rapid intestinal transit are nonspecific defenses. Humoral antibody and cell-mediated immunity play limited roles in preventing dissemination.

Epidemiology

Fecal-oral transmission of cysts involves contaminated food or water. Amebas can be transmitted directly by sexual contact involving the anus.

Diagnosis

Acute diarrhea is the usual presentation of symptomatic disease. Ulceration is associated with occult or gross blood in stool and/or with a visceral abscess. The condition may be confirmed by identification of E histolytica in the stool or in abscess aspirates. The ameba in abscesses line the wall of the abscess cavity and thus will be found in the last material aspirated from the abscess. Ameba can be cultured. Positive serologic tests, particularly tests showing rising antibody levels, may provide indirect evidence of infection.

Control

Prevention is largely a matter of personal and public hygiene. There are no effective immunizations or prophylaxis. Various drugs are used to treat different clinical syndromes.

Introduction

Amebas are unicellular organisms common in the environment: many are parasites of vertebrates and invertebrates. Relatively few species inhabit the human intestine and only Entamoebahistolytica is identified as a human intestinal pathogen. A second pathogen of the human colon is Dientamoeba fragilis, which looks like an ameba under the light microscope, and was classified as a true ameba for many years, but which is now identified as a flagellate.

Entamoeba Histolytica

Clinical Manifestations

Figures 79-1 and 79-2 present an overview of the life cycle of the ameba and the pathogenesis of amebic infections. Pathogenic and non-pathogenic strains of E histolytica inhabit the human digestive tract. Even pathogenic strains may live in the lumen as benign commensals. If mucosal invasion occurs, it may be limited to a few simple superficial erosions or it may progress to total involvement of the colonic mucosa with ulceration. Table 79-1 presents a World Health Organization classification of the clinical syndromes and related pathophysiologic mechanisms of E histolytica infections. The clinical manifestations vary with the extent of involvement. Mucosal erosion causes diarrhea, which increases in severity with increasing area and depth of involvement. Symptoms are also affected by the site of the infection. The more distal the lesion in the colon, the greater the likelihood and severity of symptoms; thus small rectal lesions are more likely to be symptomatic than larger cecal lesions. Rectal bleeding is only slightly less common than diarrhea and is usually, but not invariably, associated with diarrhea. Such bleeding may be grossly apparent or may be occult and demonstrable only by chemical testing for blood. Urgency, tenesmus, cramping abdominal pain and tenderness may be present.

Figure 79-1. Pathogenesis of E histolytica infection.

Figure 79-1

Pathogenesis of E histolytica infection.

Figure 79-2. Multiplication and life cycle of E histolytica.

Figure 79-2

Multiplication and life cycle of E histolytica.

Table 79-1. Classification of Amebiasis.

Table 79-1

Classification of Amebiasis.

The intestinal syndromes caused by E histolytica form a continuum ranging in severity from mild diarrhea to hemorrhagic dysentery. The span from mild to severe diarrhea is classified as non-dysentery colitis (Table 79-1). Amebic dysentery has a dramatically different clinical presentation. The diarrhea is replaced by dysenteric stools consisting largely of pus and blood without feces. There is evidence of systemic toxicity with fever, dehydration, and electrolyte abnormalities. Tenesmus and abdominal tenderness are regular features. This fulminant presentation may occur suddenly or evolve from less severe, pre-existing disease.

Occasionally, and for no apparent reason, colonic infection with E histolytica will evoke a proliferative granulomatous response at an ulcer site. This infectious pseudotumor, called an ameboma, may become the leading point of an intussusception or may cause intestinal obstruction. This complication is uncommon.

Peritonitis as a result of perforation has been reported in connection with severe amebic colitis and, much less often, in patients with few or no symptoms. Other complications of intestinal amebiasis include colocutaneous fistula, perianal ulceration, urogenital infection, colonic stricture, intussusception, and hemorrhage. Most of these complications are uncommon and therefore may prove difficult to diagnose. The term post-amebic colitis is used for nonspecific colitis following a bout of severe acute amebic colitis. In such cases, the colon is free of parasites and the clinical findings resemble those of chronic ulcerative colitis.

Extraintestinal amebiasis begins with hepatic involvement. Many patients with acute intestinal infection also have hepatomegaly, but in these cases amebas are not demonstrable in the liver and the pathogenesis of this hepatomegaly is not clear. A focal amebic abscess in the liver represents metastasis from intestinal infection. Symptomatic intestinal infection need not be present. The abscess appears as a slowly enlarging liver mass. Often the patient will have right upper quadrant pain, which may be referred to the right shoulder. If the abscess is located in a palpable portion of the liver, the area will be tender. Occasionally the enlarging abscess presses on the common bile duct and causes jaundice. If located under the dome of the diaphragm, the abscess may cause elevation of the dome of the diaphragm which presses on the right lung base, causing atelectasis and physical findings of consolidation. As the abscess nears the diaphragm the inflammation may stimulate pleural effusion.

Pleural, pulmonary, and pericardial infection occurs as a result of direct extension from the liver. Lung involvement is far more common than pericardial infection. Infection metastatic from the liver can involve other viscera or can give rise to a brain abscess. However, these complications are uncommon.

Structure

E histolytica has a relatively simple life cycle that alternates between trophozoite and cyst stages (Figs. 79-1 and 79-2). The trophozoite is the actively metabolizing, mobile stage, and the cyst is dormant and environmentally resistant. Diagnostic concern centers on both stages (Fig. 79-3 and Table 79-2). Trophozoites vary remarkably in size-from 10 to 60 µm or more in diameter, and when they are alive they may be actively motile. Amebas are anaerobic organisms and do not have mitochondria. The finely granular endoplasm contains the nucleus and food vacuoles, which in turn may contain bacteria or red blood cells. The parasite is sheathed by a clear outer ectoplasm. Nuclear morphology is best seen in permanent stained preparations. The nucleus has a distinctive central karyosome and a rim of finely beaded chromatin lining the nuclear membrane.

Figure 79-3. Amebas found in stool specimens of humans.

Figure 79-3

Amebas found in stool specimens of humans. (Modified from Brooke, MM, Melvin DM: Morphology of diagnostic stages of intestinal parasites of man. Public Health Service Publication No. 1966, 1969.)

Table 79-2. Intestinal Amebas of Humans.

Table 79-2

Intestinal Amebas of Humans.

The cyst is a spherical structure, 10-20 µm in diameter, with a thin transparent wall. Fully mature cysts contain four nuclei with the characteristic amebic morphology. Rod-like structures (chromatoidal bars) are present variably, but are more common in immature cysts. Inclusions in the form of glycogen masses also may be present. A number of non-pathogenic amebae can parasitize the human gastrointestinal tract and may cause diagnostic confusion. These include Entamoeba hartmanni, Entamoeba gingivalis, Entamoeba coli, Endolimax nana, and Iodamoeba butschlii (Fig. 79-3 and Table 79-2)

Classification

Many infections with E histolytica occur without evidence of invasion of the intestinal lining. Virulence in the ameba—the ability to produce intestinal invasion or extraintestinal disease—is a heritable characteristic. Morphologically identical amebas may be identified as pathogenic or non-pathogenic on the basis of size, cultural characteristics, virulence in a rat model or in tissue culture, selective agglutination by lectins, reaction with monoclonal antibodies, or isoenzyme patterns. A pathogen-specific galactose adhesion epitope is described. Ribosomal RNA sequence analysis and restriction fragment length polymorphism analysis also can separate pathogenic from non-pathogenic strains.

A number of non-pathogenic but apparently genuine E histolytic strains have been isolated from human carriers. These amebas can be cultured at room temperature as well at 37° C and will grow in hypotonic media, whereas pathogenic amebas require isotonic media and 37° C for growth. These low-temperature strains have isoenzyme patterns identical with the sewage- associated, non-pathogenic Entamoeba moshkovskii. Two classic tests to identify pathogenic strains are the ability to cause cecal ulceration in weanling rats and agglutination by the lectin concanavalin A. These tests of virulence have been supplanted by isoenzyme analysis and the use of monoclonal antibodies to identify pathogenic strains of E histolytica, but the clinical applicability of this technique is pending.

Isoenzyme patterns are known for four amebic enzymes: glucose phosphate isomerase (GPI), hexokinase (HK), malate:NADP+ oxidoreductase (ME), and phosphoglucomutase (PGM). The isoenzyme patterns of three of these, GPI, HK, and PGM, can be used to define 20 zymodemes of E histolytica. The enzyme markers associated with pathogenicity are the presence of a b band and the absence of an a band for PGM.

Zymodemes II, VI, VII, XI, XII, XIII, XIV, XIX, and XX are pathogenic. Zymodemes II and XI are responsible for liver abscesses. There have been several reports of cultured amebas undergoing a change in zymodeme pattern after manipulation of associated bacterial flora. Attempts to reproduce these observations have not been successful. Zymodeme patterns are of epidemiologic and research interest but their limited availability makes them less useful clinically. A number of other factors, primarily environmental, that affect virulence are discussed below.

It is possible to distinguish with monoclonal antibodies the galactose-specific adhesions from pathogenic and non-pathogenic ameba. This offers the possibility of simplified laboratory determination of pathogenicity.

Multiplication and Life Cycle

Amebas multiply in the host by simple binary fission. Most multiplication occurs in the host, and survival outside the host depends on the desiccation-resistant cyst form. Encystment occurs apparently in response to desiccation as the ameba is carried through the colon. After encystment, the nucleus divides twice to produce a quadrinucleate mature cyst. Excystment occurs after ingestion and is followed by rapid cell division to produce four amebas which undergo a second division. Each cyst thus yields eight tiny amebas.

Pathogenesis

The fecal-oral transmission of the ameba usually involves contaminated food or water. The parasite can also be transmitted directly by ano-genital or oro-anal sexual contact. Latent infections can become invasive in a setting of impaired host immunity.

Ingested cysts of E histolytica excyst in the small intestine (Figs 79-1 and 79-2). Trophozoites are carried to the colon, where they mature and reproduce. The parasite may lead a commensal existence on the mucosal surface and in the crypts of the colon. Successful colonization depends on factors such as inoculum size, intestinal motility, transit time, the presence or absence of specific intestinal flora, the host's diet and the ability of the ameba to adhere to the colonic mucosal cells. The ameba adherence molecule has been identified as a lectin which can bind to either of two common carbohydrate components of cell membrane, galactose and N-acetyl glactoseamine. Binding to colonic mucins blocks adherence to mucosal cells. Depletion of mucus results in binding to the mucosa, an essential step in the development of the disease. If amebas pass down the colon they encyst under the stimulus of desiccation, and then are evacuated with the stool.

The factors that lead to tissue invasion by E histolytica are poorly understood. The genetic virulence factors mentioned above play a major role but several environmental factors are also important. Although the mechanisms of action are not clear, both changes in the intestinal flora and the nature of the host's diet have been implicated. All virulence factors come to a final common path where the ameba attacks and kills the host cell. Binding involving the galactose adherence lectin is essential for the cytolytic effect. Blocking adherence blocks the killing. This cytolytic event is a result of incorporation in the host cell membrane of an ameba-produced, pore-forming protein, amoebapore. This protein forms ion channels in lipid cell membranes and results in cell death within minutes of cell contact with the ameba. Amoebapore has been isolated, synthesized and well characterized. Non-pathogenic strains of E. histolytica can also produce amoebapore but are much less efficient at its production and the molecule is not exactly similar to that produced by virulent strains.

The initial lesion is in colonic mucosa, most often in the cecum or sigmoid colon. The slow transit of the intestinal contents in these two locations seems an important factor in invasion of the mucosa, both because it affords the ameba greater mucosal contact time and because it permits changes in the intestinal milieu that may facilitate invasion. The initial superficial ulcer may deepen into the submucosa and muscularis to become the characteristic flask-shaped, chronic amebic ulcer. Spread may occur by direct extension, by undermining of the surrounding mucosa until it sloughs, or by penetration that can lead to perforation or fistulous communication to other organs or the skin. If the amebas gain access to the vascular or lymphatic circulation, metastases may occur first to the liver and then by direct extension or further metastasis to other organs, including the brain.

Virulent E histolytica strains are capable of penetrating intact intestinal mucosa. The infection is not opportunistic and does not require pre-existing mucosal damage. Numerous proteases have been isolated in E histolytica; however, the mechanism of penetration remains unclear. Metastatic foci present as abscesses with a central zone of lytic necrosis surrounded by a zone of inflammatory cell infiltration. Metastatic abscesses behave as space-occupying lesions unless they become secondarily infected or rupture.

The clinical presentation of intestinal infections depends on the extent and anatomic location of the ulceration and mucosal damage. Small, sparse ulcerations may be asymptomatic. As the involved area of the mucosa increase in size and/or in depth, motility disturbances occur, primarily diarrhea with cramping pain. Exudation from the denuded mucosa adds to intestinal content. When the mucosal involvement becomes extensive, diarrhea is replaced by dysentery, with the passage of exudate, blood and mucus. Toxic megacolon and perforation are rare complications of extensive involvement. Systemic signs of infection include fever, rigor, and polymorphonuclear leukocytosis.

Host Defenses

The gastric acid “barrier” and the steady movement of food through the intestine are nonspecific defense mechanisms invoked to explain both the experimental observation that large inocula are required to produce consistent infection in animals and the pathologic observation that few lesions are found in the small intestine, a zone of rapid transit. A role for colonic mucins in protection and depletion of these mucins in infection has been suggested.

Usually amebas alone stimulate little or no direct cellular response. Primary intestinal lesions elicit little reaction until secondary bacterial infection occurs. Amebic abscesses similarly elicit only a mild leukocytic response, which may be largely a response to the host cellular debris in the abscess. Amebas are antigenic and stimulate an antibody response and cellular sensitivity. In vivo studies have yielded contradictory results regarding the response of amebas to exposure to humoral antibodies. The occurrence of progressive and/or recurrent infection in the face of established immune sensitivity suggests that the host immune response is relatively ineffective against established infections.

Epidemiology

Fecal-oral transmission occurs when food preparation is not sanitary or when drinking water is contaminated. Contamination may come directly from infected food handlers or indirectly from faulty sewage disposal. Endemic or epidemic disease may result. The prevalence of amebiasis in underdeveloped countries reflects the lack of adequate sanitary systems.

Amebas are found in all climates, arctic to tropical. Symptomatic infections (amebic disease) are far more prevalent in certain geographic foci, and this uneven prevalence of disease, as opposed to infection, is now explained by the variable geographic predominance of pathogenic zymodemes. Similar environments thus are likely to have a comparable infection rate but may have a widely different disease prevalence.

Diagnosis

Table 79-1 gives the classification of the clinical syndromes caused by E histolytica, adopted by the World Health Organization, and their related pathophysiologic mechanisms. Amebic infections are diagnosed definitively by identifying the ameba in stool or exudate (see Fig. 79-4). Under some circumstances, however, the physician must settle for a presumptive diagnosis based on serologic or clinical evidence alone. Diagnosis may be difficult if few organisms are shed in the stool. Effective methods exist for concentrating cysts but not trophozoites in stool specimens. Fortunately, a direct relationship is usually seen (although there are exceptions) between the severity of disease and the number of amebas shed in the stool; hence, the more severe the infection the easier the diagnosis. Unfortunately, a number of substances that may be administered to the patient in the course of diagnosis or therapy can impair the ability to make a direct diagnosis. These compounds can suppress the shedding of amebas into the stool but may not interfere with the course of invasion infection. Such compounds include barium, bismuth, kaolin, soapsuds (as enemas), and antimicrobials that can reach the intestinal lumen. The suppression of shedding may be short-lived (soapsuds enema), or may last weeks or months (broad-spectrum antibiotics). These compounds render timely direct diagnosis unreliable and often impossible.

Figure 79-4. Evaluation of suspected cases of intestinal amebiasis.

Figure 79-4

Evaluation of suspected cases of intestinal amebiasis.

Amebas may be identified in direct smears, but specific diagnosis usually depends upon obtaining a fixed stained preparation. Trophozoites deteriorate rapidly in stool specimens, and therefore preservatives, either polyvinyl alcohol or the merthiolate-iodine-formaldehyde (MIF) combination, are important diagnostic aids. Finally, it is unrewarding to search for trophozoites in formed stool because most trophozoites encyst as the stool desiccates.

Trophozoites can be found in diarrhea. Most infections in formed stool specimens will be detected by examining three specimens passed over a 7- to 10-day period. A negative examination of single stool specimen does not rule out infection. Trophozoites may be obtained by administering a purgative agent or by scraping suspicious lesions at the time of sigmoidoscopy.

Amebas are difficult to demonstrate in aspirates from extraintestinal abscesses (Fig. 79-5) unless special precautions are taken. The contents of most amebic abscesses are relatively free of the organism. Instead, the organisms concentrate adjacent to the wall of the abscess cavity. If care is taken during aspiration to separate serial aliquots of aspirate, amebas may be found in the last syringe that empties the cavity. Cysts or trophozoites are only found in approximately one-half of all patients with amebic liver abscess.

Figure 79-5. Evaluation of suspected cases of hepatic amebiasis.

Figure 79-5

Evaluation of suspected cases of hepatic amebiasis.

Serologic studies (Fig. 79-5) may be useful, particularly when direct diagnosis is not possible. Such methods include gel diffusion, immunoelectrophoresis, countercurrent electrophoresis, indirect hemagglutination, indirect fluorescent antibody, skin tests, enzyme-linked immunosorbent assay (ELISA) and latex agglutination. Many of these techniques are best suited for immunoepidimiology, but gel diffusion, countercurrent electrophoresis, and latex agglutination are available for clinical studies because they are readily run on a single serum sample. A positive result on these tests indicates only prior experiences with invasive amebiasis. In environments where the incidence of amebiasis is low, such as in the United States, a positive antibody test often indicates active disease, an impression strengthened if the clinical findings agree. In areas of high prevalence a single positive antibody test is less significant. The physician rarely observes the patient long enough to measure a rising titer as evidence of active ongoing invasive infection.

Amebas may be cultured from the stool. However, because the techniques involved are somewhat more cumbersome than those routinely used for bacterial organisms, culturing is not widely used as a diagnostic tool. It is essential for virulence testing.

Testing with monoclonal antibodies demonstrates ameba in the stool, and, if the galactose adhesion epitopes are tested for, pathogenicity may be determined as well. Broad scale application is proposed.

A number of nonpathogenic amebas that can inhabit the human intestinal tract may confuse direct diagnosis. These include Entamoeba hartmanni, Entamoeba gingivalis, Entamoeba coli, Endolimax nana, and Iodamoeba butschlii. Although these parasites do not cause illness, they indicate that the patient has ingested feces-contaminated food or water, so their presence may prompt careful study of additional specimens (Fig. 79-3).

Control

Preventive measures are limited to environmental and personal hygiene. Treatment depends on drug therapy, which in the case of some abscesses must be supplemented with drainage, either open or by aspiration. Effective drugs are available for liver abscess but intestinal infection is less successfully treated. No single drug is completely effective in eradicating amebas from the gut, so reliance is often placed on combination therapy.

Acute intestinal disease is best treated with metronidazole at a dose of 750 mg three times a day orally for 10 day. In children the dose is 40 mg/kg/day divided into three doses and given orally for 10 days. While this treatment is effective against invasive intestinal disease, it is less effective in clearing amebas from the intestine. Patients unable to take metronidazole may be given a broad spectrum antibiotic for two weeks. It too is relatively ineffective at clearing the amebas from the gut. There are two choices for a drug to clear amebas from the lumen of the gut: iodoquinol at an adult dose of 650 mg orally three times daily for 20 days or diloxanide furoate at an adult dose of 500 mg orally three times daily for 10 days.

Amebic liver abscess is best treated with metronidazole at several possible dose regimens, but cases of drug failure have been reported. Chloroquine or dehydroemetine are less desirable alternatives. Aspiration of the abscess is not helpful except for diagnostic purposes unless rupture is imminent. Amebic abscesses heal at the same rate with or without aspiration. Abscesses with secondary bacterial infection must be drained surgically. Abscesses involving other organs respond less well to drugs and require drainage.

References

  1. Beaver PC, Jung RC, Cupp EW: Clinical Parasitology. 9 Ed. Lea & Febiger, Philadelphia, 1984 .
  2. Haque R, Kress K, Wood S. et al. Diagnosis of pathogenic Entamoeba histolytica infection using a stool ELISA based on monoclonal antibodies to the galactose-specific adhesion. J Inf Dis. 1993;167:247. [PubMed: 8369019]
  3. Jackson TFHG, Cathiram V, Simjee AE. Seroepidemiological study of antibody responses to the zymodemes of Entamoeba histolytica . Lancet. 1985;1:716. [PubMed: 2857995]
  4. Kean BH. The treatment of amebiasis. JAMA. 1976;235:501. [PubMed: 1107608]
  5. Leippe M, Muller-Eberhard H. The pore-forming peptide of Entamoeba histolytica, the protozoan parasite causing human amoebiasis. Toxicology 87. 1994;5:18. [PubMed: 7512763]
  6. Martinez-Palomo A, Gonzalez-Robles A, de la Torre M. Selective agglutination of pathogenic strains of Entamoeba histolytica induced by Con A. Nature. 1973;245:186. [PubMed: 4355103]
  7. Miller V, Spencer MJ, Chapin M. et al. Dientamoeba fragilis, a protozoan parasite in adult members of a semicommunal group. Dig Dis Sci. 1983;28:335. [PubMed: 6403320]
  8. Mirelman D, Bracha R, Wexler A. et al. Entamoeba histolytica: Effect of growth conditions and bacterial associates on isoenzyme patterns and virulence. Exp Parasitol. 1986;62:142. [PubMed: 2873049]
  9. Mirelman D, Bracha R, Wexler A. et al. Changes in isoenzyme patterns of a cloned culture of nonpathogenic Entamoeba histolytica during axenization. Infect Immun. 1986;54:827. [PMC free article: PMC260244] [PubMed: 2877951]
  10. Petri W, Mann B. Molecular mechanisms of invasion by Entamoeba histolytica. Cell Biology 4. 1993;305:313. [PubMed: 7504959]
  11. Sargeaunt PG, Baveja UK, Nanda R. et al. Influence of geographical factors in the distribution of pathogenic zymodemes of Entamoeba histolytica: Identification of zymodeme XIV in India. Trans Soc Trop Med Hyg. 1984;78:96. [PubMed: 6324421]
  12. Sargeaunt PG, Williams JE, Neal RA. A comparative study of Entamoeba histolytica (NIH:200, HK9, etc) “E histolytica-like” and other morphologically identical amoebae using isoenzyme electrophoresis. Trans R Soc Trop Med Hgy. 1980;74:469. [PubMed: 6255635]
  13. Spice W, Cruz-Reyes J, Ackers J: Entamoeba histolytica. Molecular and Cell Biology of Opportunistic Infections in AIDS, Chapman & Hall, London, 1992.
  14. Strachan WD, Spice WM, Chiodini PL. et al. Immunological differentiation of pathogenic and non-pathogenic isolates of Entamoeba histolytica. Lancet. 1988;1:561. [PubMed: 2894495]
  15. World Health Organization: Amoebiasis: Report of a WHO Expert Committee. WHO Tech Rep Ser 1969 . [PubMed: 4978968]
Copyright © 1996, The University of Texas Medical Branch at Galveston.
Bookshelf ID: NBK7742PMID: 21413264

Views

  • PubReader
  • Print View
  • Cite this Page

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Similar articles in PubMed

See reviews...See all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...