Prospective Study of the Prevalence, Genotyping, and Clinical Relevance of Dientamoeba fragilis Infections in an Australian Population
http://www.100md.com
微生物临床杂志 2005年第6期
St. Vincent's Hospital, Department of Microbiology, Sydney, Australia
University of Technology Sydney, Institute for the Biotechnology of Infectious Diseases, St. Leonard's Campus, Sydney, Australia
University of Technology Sydney, Department of Cell and Molecular Biology, St. Leonard's Campus, Sydney, Australia
ABSTRACT
A prospective study was conducted over a 30-month period, in which fecal specimens from 6,750 patients were submitted to the Department of Microbiology at St. Vincent's Hospital, Sydney, Australia. Trophozoites of Dientamoeba fragilis were detected in 60 (0.9%) patients by permanent staining, and confirmation was performed by PCR. Gastrointestinal symptoms were present in all patients, with diarrhea and abdominal pain the most common symptoms. Thirty-two percent of patients presented with chronic symptoms. The average age of infected patients was 39.8 years. No correlation was found between D. fragilis and Enterobius vermicularis, a proposed vector of transmission for D. fragilis. The genetic diversity of 50 D. fragilis isolates was examined by PCR, and the PCR products were analyzed for the presence of restriction fragment length polymorphisms. These results showed no variation in the small-subunit rRNA gene and demonstrated a single genotype for all Australian isolates. This study shows the potential pathogenic properties of D. fragilis and the need for all laboratories to routinely test for this organism.
INTRODUCTION
Dientamoeba fragilis is a trichomonad parasite found in the gastrointestinal tract of humans and implicated as a cause of gastrointestinal disease. Dientamoeba fragilis has been found in most parts of the world in both rural and cosmopolitan areas (10). The prevalence of this organism in Australia varies greatly, from 0.4% to 16.8%, in patients presenting with gastrointestinal complaints (1, 22).
No cyst stage has been observed, and only the trophozoites are detected in stool samples. Definitive diagnosis is based on prompt fixation and permanent staining, as the trophozoites degenerate rapidly, within hours of been passed, and demonstration of their characteristic nuclear structure cannot be achieved in unstained preparations (24). Daily shedding of D. fragilis trophozoites has been shown to be highly variable, with intermittent shedding occurring regularly, necessitating multiple sampling for maximum chances of detection (20).
Molecular techniques for the diagnosis of D. fragilis show much promise, with PCR demonstrating excellent sensitivity and specificity (18). Such techniques have been used successfully for the diagnosis of other pathogenic protozoa (11, 19).
Molecular genotyping and sequence analysis have demonstrated that D. fragilis exists as two genetically distinct forms (9, 15, 18, 26). Stark et al. (18) sequenced the SSU rRNA gene of seven Australian D. fragilis isolates, and the data generated from the seven showed no variation among them. These observations support the notion that D. fragilis is a clonal species. The sequences from the Australian isolates, however, differed from the sequence of the D. fragilis strain Bi/PA (ATCC 30948; GenBank accession no. U37461) and were found to be similar to those found in a recent study in the Netherlands (15). The true incidence of the wild-type and variant forms in Australia needs to be established and to determine if such variation has any influence on the pathogenicity of the parasite.
A prospective study was undertaken to determine the prevalence and clinical relevance of D. fragilis infections in an Australian population and to determine the genetic diversity of these isolates obtained at the small-subunit (SSU) rRNA gene locus.
MATERIALS AND METHODS
Fecal specimens. All fecal specimens submitted to the Department of Microbiology at St. Vincent's Hospital, Sydney, for investigation of diarrhea from March 2002 until July 2004 were included in the study. One to three specimens per patient were examined. Specimens from outpatients were collected by the patient and submitted to the laboratory as a fresh specimen along with a portion mixed with sodium acetate-acetic acid-formalin (SAF) preservative. Specimens from inpatients or received without a portion fixed in SAF were immediately preserved in SAF upon arrival at the laboratory.
Microbiological investigation. Fecal specimens were cultured for the following bacterial pathogens: Salmonella spp., Shigella spp., Campylobacter spp., Aeromonas spp., Yersinia spp., and Clostridium difficile, and culture for Vibrio spp. was performed where indicated if a patient had history of travel to an endemic area, using standard laboratory procedures and techniques.
An immunochromatographic screening test, the Adeno/Rota STAT-PAK (Chembio Diagnostic Systems Inc., Sydney) for the detection of adenovirus and rotavirus antigen in feces was used according to the manufacturer's recommendations.
Approximately 1 g of feces was placed into SAF and fixed overnight. The fixed specimens were then stained using a modified iron hematoxylin stain, incorporating a carbol fuchsin step to detect coccidia (Fronine, Australia), according to the manufacturer's recommendations. Formalin-ethyl acetate concentration was used for the detection of any helminth ova. In addition, any specimens from human immunodeficiency virus-infected patients were examined for microsporidial spores using the Uvitex 2B stain (21).
PCR for D. fragilis. All specimens where D. fragilis was detected by permanent stain underwent DNA extraction and PCR for D. fragilis-specific DNA using primers DF400 and DF1250 as previously described (18).
Restriction fragment length polymorphism. Restriction fragment length polymorphism (RFLP) analysis was undertaken on all positive PCR products. Eight μl of the PCR product was digested with 10 U of DdeI (Roche, Australia) in a final volume of 15 μl for 1 h at 37°C. Samples were analyzed by electrophoresis on 3% ReadyAgarose gels (Bio-Rad, Sydney).
Follow-up data. Clinical data were collected from all patients diagnosed with D. fragilis. Wherever possible multiple, sticky-tape tests, two to five tapes per patient, were conducted for the detection of Enterobius vermicularis.
Control group. A control group comprising 900 fecal samples from patients without diarrhea or symptoms of gastroenteritis (submitted for occult blood testing and fecal reducing substances) were used. These specimens were processed as above and stained using a modified iron hematoxylin stain. Ninety of these specimens were randomly selected and underwent PCR using D. fragilis-specific primers as described by Stark et al. (18).
Questionnaire. Questionnaires were distributed to 26 laboratories in the Sydney metropolitan area. Information requested for the calendar years 1996 to 2002 included total number of fecal samples processed for ova cysts and parasites, total number of specimens positive for D. fragilis, use of permanent stain, fixation method used in this period, and the situation in which a fixation method would be used.
RESULTS
A total of 6,750 patients submitted fecal specimens between March 2002 and July 2004. Sixty patients were diagnosed with D. fragilis infection from the permanent stains, giving a prevalence of infection of 0. 9%. The results found in this study are summarized in Table 1.
Of the 60 patients infected with D. fragilis, six (10%) had a history of recent overseas travel; three to Southeast Asia, one to Timor, one to Fiji, and one to Papua New Guinea. The remaining 54 of 60 patients (90%) had no recent history of travel outside Australia.
A total of 24/60 (40%) patients had other parasites detected (Table 2). No coccidian parasites were detected. The only other pathogenic protozoan was Giardia intestinalis, which was found concurrently with D. fragilis in three samples. The remaining 36 patients (60%) had only D. fragilis detected. All fecal samples were semiformed or liquid.
The most frequent clinical symptoms associated with D. fragilis infection were diarrhea, abdominal pain, and loose bowel movements. Vomiting was only reported in one patient. Chronic persistent symptoms were common, with 19/60 (32%) patients having diarrhea of over 2 weeks in duration, and one patient claimed to have intermittent diarrhea for several years. Five patients had recurrent D. fragilis infections. One patient was diagnosed with irritable bowel syndrome. All patients were symptomatic. Only one patient (human immunodeficiency virus infected) was immunosuppressed, with all the others being immunocompetent. No microsporidia were detected in the human immunodeficiency virus-infected patient.
Thirty patients were female and 30 were male, with the age range being 3 to 79 years (Fig. 1). The average age was 39.8 years, with a median of 44.5 years. No seasonal variation was found with D. fragilis infection.
No helminth ova were detected in the 60 patients using a formalin-ethyl acetate concentration technique, and no Enterobius vermicularis adults or ova were found; 33/60 (55%) patients submitted a sticky-tape test for E. vermicularis ova, all of which were negative.
No bacterial pathogens were isolated from the patients with D. fragilis infection. The immunochromatographic tests for both adenovirus and rotavirus were also negative for all of the patients.
PCR was performed on 54 of the 60 samples; for six specimens there was a delay (>7 days) in undertaking the DNA extraction, so these specimens were excluded from PCR testing. A specific D. fragilis PCR product of approximately 870 bp was detected in 50 out of 54 samples using the D. fragilis-specific primers designed by Stark et al. (18). RFLP was performed on the 50 positive PCR samples. All gave identical RFLP patterns (data not shown).
Nine hundred fecal samples from patients without gastrointestinal symptoms were used as a control group. No D. fragilis was detected by permanent staining. However, nonpathogenic protozoa were detected in the control group. Blastocystis hominis was found in 47 (5.2%) patients and Endolimax nana in 19 (2.1%), while Blastocystis hominis and Endolimax nana were found concurrently in 12 (1.3%) patients. One patient (0.1%) was found to have Entamoeba hartmanni. PCR using D. fragilis-specific primers was undertaken on 90 samples randomly chosen from the control group. All 90 specimens were negative for D. fragilis DNA by PCR.
Of the 26 laboratories that were sent the questionnaire, only 11 responded. The remaining 15 laboratories were contacted, and four agreed to participate in a phone interview using the same questions as on the written questionnaire. Of the 15 laboratories it was determined that only three in the Sydney metropolitan area routinely performed permanent stains on feces for ova, cysts, and parasite examinations.
DISCUSSION
Dientamoeba fragilis has a worldwide cosmopolitan distribution. In Australia and New Zealand, the reported prevalence rate ranges from 0.4% in western Australia (1) and 1.5% in an urban community in Brisbane (17) to 2.2% in Christchurch, New Zealand (14), and 16.8% in suburban Sydney (21). A longitudinal study of parasite infections in Aboriginal children from the Queensland outback found a prevalence of 5.0% for D. fragilis (23). In this present study a prevalence of 0.9% was found; this is in stark contrast to the prevalence of 16.8% that was found by Walker et al. (22) in the Sydney suburb of French's Forrest.
In this study, D. fragilis infection was closely associated with diarrhea, abdominal pain, and loose bowel movements. All patients with D. fragilis infection were symptomatic, and bacterial and viral causes of these symptoms are unlikely, as routine microbiological cultures and adenovirus and rotavirus testing were negative. However, testing for pathogenic Escherichia coli or norovirus was not undertaken. Three patients were also infected with Giardia intestinalis, which could have caused the gastrointestinal symptoms described in those patients.
One important finding of this study was that chronic persistent symptoms were common. Thirty-two percent of patients had diarrhea for more than 2 weeks, and one patient claimed to have had intermittent diarrhea for several years. Five patients had recurrence of symptoms during the course of the study. It is unknown whether these recurrences were due to treatment failure or reinfection from a common source. One patient was diagnosed with irritable bowel syndrome and was subsequently found to have D. fragilis infection. A recent Australian study by Borody et al. (2) showed a link between D. fragilis and irritable bowel syndrome. Twenty-one patients diagnosed with irritable bowel syndrome and concurrent D. fragilis infection were treated with iodoquinol and doxycycline. Complete elimination of D. fragilis with marked clinical improvement occurred in the majority of patients.
Ten percent of patients diagnosed with D. fragilis infection had a history of recent overseas travel, including Southeast Asia, Papua New Guinea, Timor, and Fiji. Dientamoeba fragilis has been implicated as a cause of diarrhea in returning Swedish travelers, with Norberg et al. (13) finding 63% of patients in a retrospective study had been infected outside the country. Most patients were infected in Africa, South America, and the Middle East.
No parasites were detected by formalin-ethyl acetate concentrations performed on fecal specimens from the D. fragilis-infected patients. Fifty-five percent of the patients submitted multiple tape tests for the detection of Enterobius vermicularis ova, and no E. vermicularis ova were detected. Many researchers have postulated that pinworm is a vector for D. fragilis transmission. Burrows and Swerdlow (3) were the first to propose that E. vermicularis might be a vector for D. fragilis. Several other researchers also found a higher than expected concurrence of D. fragilis and E. vermicularis coinfections (4, 5, 16, 26).
In contrast, a recent study of 25 pediatric cases of D. fragilis found no infections were associated with E. vermicularis (7). These results, along with the findings from this present study, would argue against the hypothesis that E. vermicularis plays a significant role in the transmission of D. fragilis. Most studies that have examined D. fragilis infection have inadequately examined for E. vermicularis. It has yet to be proven what role helminth ova play in the transmission of D. fragilis. Further study is required to ascertain the true mode of transmission of this organism.
Other enteric protozoa were present in 40% of patients with D. fragilis infection. The most common organism was B. hominis. Other protozoa present included E. nana, E. hominis, E. coli, Iodamoeba butschlii, and G. intestinalis. All of these parasites are known to be transmitted via the fecal-oral route. Other researchers have found similar rates of coinfection of D. fragilis with other parasites that are transmitted via the fecal-oral route. Windsor et al. (25) found 54% of patients with D. fragilis had other parasites or enteropathogens present. These findings provide circumstantial evidence to support a hypothesis for a fecal-oral route of transmission for D. fragilis.
No D. fragilis trophozoites were detected in the control group of 900 smaples from patients without gastrointestinal symptoms. This is in contrast to other studies, where D. fragilis was detected in patients with no clinical symptoms (6) and in a case-control study on gastroenteritis from the Netherlands, where D. fragilis was recovered more frequently from controls than case patients (8). These findings may be attributed to the fact that asymptomatic carriage of intestinal protozoa can often occur.
The permanent stained smears positive for D. fragilis were confirmed by PCR. A sensitivity of 93% (50/54 samples) was obtained using a previously published method (18). All 90 negative samples from the control group failed to produce a PCR product.
Sequence data generated in several studies supports the notion for at least two distinct genetic variants of D. fragilis. Johnson and Clarke (9) estimated a sequence divergence of 2% between the two SSU rRNA genotypes of D. fragilis; this was later supported by Peek et al. (15) by sequencing a 558bp region of the SSU rDNA. Sequence data generated by Stark et al. (18) from the entire SSU rDNA region of Australian isolates of D. fragilis showed a greater sequence divergence of 4% between the Australian genotypes and the D. fragilis strain Bi/PA (ATCC 30948). All Australian strains sequenced were identical, which supports the notion that D. fragilis is a clonal species.
The Australian isolates were found to be similar to those found in a recent study in the Netherlands and do not contain the polymorphic DdeI restriction site (CTTAG) at position 644 found in D. fragilis strain Bi/PA (15). RFLP analysis was undertaken on all 50 Australian samples to determine the genotypes present in the Australian population and the extent of genetic diversity. The PCR used in this study amplifies the SSU rRNA region from approximately position 400 to position 1270. This PCR product contains two DdeI restriction sites (CTTAG) that are present in the D. fragilis ATCC 30948 strain yet are absent in the Australian genotypes. All 50 D. fragilis samples showed no variation and corresponded to genotype A. These findings suggest that D. fragilis in Sydney, Australia, displays only a single genotype in fecal samples from various groups, including inpatients, outpatients, and travelers. Further studies are needed to identify the presence of other genotypes throughout Australia.
Dientamoeba fragilis has no recognized cyst stage, and as such, diagnosis is dependent on detecting the trophozoites. As these trophozoites degenerate rapidly prompt fixation of the specimen is necessary (26). Successful diagnosis of D. fragilis is closely associated with the use of permanent stains of fecal smears. Failure to use permanent staining and fixation techniques will inevitably preclude identification of D. fragilis. The aim of the questionnaire sent to the Sydney laboratories was to determine how many laboratories routinely undertake permanent staining and therefore how many laboratories are able to report the presence of D. fragilis. Of the 26 Sydney laboratories, 58% (15/26) participated in the survey, and only three routinely performed permanent staining for ova, cyst and parasites on fecal specimens. Those three laboratories were the only institutions that detected D. fragilis in routine samples. Therefore the true extent of D. fragilis infection must be greatly underestimated as most laboratories do not use techniques to adequately identify this organism.
This is the first prospective study of D. fragilis in Australia to examine clinical data in addition to the genetic diversity of the isolates. Diagnosis was based on permanent staining of fixed fecal smears and confirmed by PCR which demonstrated good sensitivity. All patients infected with D. fragilis were symptomatic and D. fragilis infections were most commonly associated with diarrhea and abdominal pain. Concurrent infections with other protozoa were common, occurring in 40% of samples. The occurrence of D. fragilis with other protozoa that are transmitted via the fecal-oral route would strengthen the case for D. fragilis also being transmitted via this route. No correlation was found with E. vermicularis or any other helminths, questioning the role, if any, pinworm has in the transmission of D. fragilis.
The genetic diversity within 50 samples was examined by PCR followed by RFLP. These data indicated that a single genotype of D. fragilis was represented, one that is genetically different from the North American D. fragilis strain Bi/PA (ATCC 30948). The evidence that D. fragilis is a pathogen is overwhelming (2, 10, 13, 16, 26), and as such all laboratories should attempt to identify this protozoan by the use of permanent staining techniques or molecular methods.
ACKNOWLEDGMENTS
This work was supported by a grant from the Institute of Laboratory Medicine at St. Vincent's Hospital, Sydney, Australia.
REFERENCES
Anonymous.1992 . Western Australian enteric pathogen report.Communicable Dis. Intell. 16:154-159.
Borody, T. J., E. F. Warren, A. Wettstein, G. Robertson, P. Recabarren, A. Fontella, K. Herdnman, and R. Surace.2002 . Eradication of Dientamoeba fragilis can resolve IBS-like symptoms. J. Gastroenterol. Hepatol. 17(Suppl.):A103.
Burrows, R. B., and M. A. Swerdlow. 1956. Enterobius vermicularis as a probable vector of Dientamoeba fragilis. Am. J. Trop. Med. Hyg. 5:258-265.
Burrows, R. B., M. A. Swerdlow, J. K. Frost, and C. K. Leeper. 1954. Pathology of Dientamoeba fragilis infections in the appendix.Am. J. Trop. Med. Hyg. 3:1033-1039.
Cerva, L., M. Schrottenbaum, and V. Kliment. 1991. Intestinal parasites: a study of human appendices. Fiola Parasitol. 38:5-9.
Colea, A., R. Silard, D. Panaitescu, P. Florescu, N. Roman, and T. Capraru. 1980. Studies on Dientamoeba fragilis in Romania. II. Incidence of Dientamoeba in healthy persons. Arch. Roum. Pathol. Exp. Microbiol. 39:49-53.
Cuffari, C., L. Oligny, and E. G. Seidman. 1998. Dientamoeba fragilis masquerading as allergic colitis.J. Paediatr. Gastroenterol. Nutr. 26:16-20.
De Wit, M. A. S., M. P. G. Koopmans, L. M. Kortbeek, N. J. van Leeuween, J. Vinje, and Y. T. H. P. van Duynhoven. 2001. Etiology of gastroenteritis in sentinel general practises in the Netherlands. Clin. Infect. Dis. 33:280-288.
Johnson, J. A., and G. C. Clarke. 2001. Cryptic Diversity in Dientamoeba fragilis. J. Clin. Microbiol. 12:4653-4354.
Johnson, E. H., J. J. Windsor, and G. C. Clark. 2004. Emerging from obscurity: biological, clinical, and diagnostic aspects of Dientamoeba fragilis.Clin. Microbiol. Rev. 17:553-570.
Limor, J. R., A. A. Lal, and L. Xiao.2002 . Detection and differentiation of Cryptosporidium parasites that are pathogenic for humans by real-time PCR. J. Clin. Microbiol. 40:2335-2338.
Mendez, O. C., G. Szmulewicz, C. Menghi, S. Torres, G. Gonzalez, and C. Gatta. 1994. Comparison of intestinal parasite infestation indexes among HIV-positive and -negative populations.Medina (Buenos Aires) 54:307-310.
Norberg, A., C. E. Nord, and B. Evengard. 2003. Dientamoeba fragilis-a protozoal infection which may cause severe bowel distress. Clin. Microbiol. Infect. 9:65-68.
Oxner, R., G. P. Paltridge, B. A. Chapman, H. B. Cook, and P. F. Sheppard. 1987. Dientamoeba fragilis: a bowel pathogen N.Z. Med. J. 100:64-65.
Peek, R., F. R. Reedeker, and T. van Gool. 2004. Direct amplification and genotyping of Dientamoeba fragilis from human stool specimens. J. Clin. Microbiol. 42:631-635.
Priess, U., G. Ockert, S. Broemme, and A. Otto. 1991. On the clinical importance of Dientamoeba fragilis infections in children. J. Hyg. Epidemiol. Microbiol. Immunol. 35:27-34.
Sawangjaroen, N., R. Luke, and P. Prociv. 1993. Diagnosis by faecal culture of Dientamoeba fragilis infections in Australian patients with diarrhoea. Trans. R. Soc. Trop. Med. Hyg. 87:163-165.
Stark, D., N. Beebe, D. Marriott, J. T. Ellis, and J. Harkness.2005 . Detection of Dientamoeba fragilis in fresh stool specimens using PCR. Int. J. Parasitol. 35:57-62.
Troll, H., H. Marti, and N. Weiss. 1997. Simple differential detection of Entamoeba histolytica and Entamoeba dispar in fresh stool specimens by sodium acetate-acetic acid-formalin concentration and PCR. J. Clin. Microbiol. 35:1701-1705.
Van Gool, T., R. Weijts, E. Lommerse, and T. G. Mank.2003 . The triple faeces test: an effective tool for the detection of intestinal parasites in routine clinical practice.Eur. J. Clin. Microbiol. Infect. Dis. 22:284-290.
Van Gool, T., F. Snijders, P. Reiss, J. K. Eeftinck Schattenkerk, M. A. Van den Bergh Weerman, J. F. Bartelsman, J. J. Bruins, E. U. Canning, and J. Dankert.1993 . Diagnosis of intestinal and disseminated microsporidial infections in patients with HIV by a new rapid fluorescence technique. J. Clin. Pathol. 46:694-699.
Walker, J. C., G. Bahr, and A. S. Ehl.1985 . Gastrointestinal parasites in Sydney. Med. J. Aust. 11:143-480.
Welch, J. S., and J. E. Stuart. 1976. A longitudinal study of parasite infections in 120 Queensland Aboriginal children. Med. J. Aust. 44:14-16.
Windsor, J. J., C. G. Clark, and L. Macfarlane.2004 . Molecular typing of Dientamoeba fragilis.Br. J. Biomed. Sci. 61:152-155.
Windsor, J. J., A. M. Rafay, A. K. Shenoy, and E. H. Johnson. 1998. Incidence of Dientamoeba fragilis in faecal samples submitted for routine microbiology analysis. Br. J. Biomed. Sci. 55:172-175.
Yang, J., and T. H. Scholten. 1977. Dientamoeba fragilis: a review with notes on its epidemiology, pathogenicity, mode of transmission and diagnosis. Am. J. Trop. Med. Hyg. 26:16-22.(D. Stark, N. Beebe, D. Ma)
University of Technology Sydney, Institute for the Biotechnology of Infectious Diseases, St. Leonard's Campus, Sydney, Australia
University of Technology Sydney, Department of Cell and Molecular Biology, St. Leonard's Campus, Sydney, Australia
ABSTRACT
A prospective study was conducted over a 30-month period, in which fecal specimens from 6,750 patients were submitted to the Department of Microbiology at St. Vincent's Hospital, Sydney, Australia. Trophozoites of Dientamoeba fragilis were detected in 60 (0.9%) patients by permanent staining, and confirmation was performed by PCR. Gastrointestinal symptoms were present in all patients, with diarrhea and abdominal pain the most common symptoms. Thirty-two percent of patients presented with chronic symptoms. The average age of infected patients was 39.8 years. No correlation was found between D. fragilis and Enterobius vermicularis, a proposed vector of transmission for D. fragilis. The genetic diversity of 50 D. fragilis isolates was examined by PCR, and the PCR products were analyzed for the presence of restriction fragment length polymorphisms. These results showed no variation in the small-subunit rRNA gene and demonstrated a single genotype for all Australian isolates. This study shows the potential pathogenic properties of D. fragilis and the need for all laboratories to routinely test for this organism.
INTRODUCTION
Dientamoeba fragilis is a trichomonad parasite found in the gastrointestinal tract of humans and implicated as a cause of gastrointestinal disease. Dientamoeba fragilis has been found in most parts of the world in both rural and cosmopolitan areas (10). The prevalence of this organism in Australia varies greatly, from 0.4% to 16.8%, in patients presenting with gastrointestinal complaints (1, 22).
No cyst stage has been observed, and only the trophozoites are detected in stool samples. Definitive diagnosis is based on prompt fixation and permanent staining, as the trophozoites degenerate rapidly, within hours of been passed, and demonstration of their characteristic nuclear structure cannot be achieved in unstained preparations (24). Daily shedding of D. fragilis trophozoites has been shown to be highly variable, with intermittent shedding occurring regularly, necessitating multiple sampling for maximum chances of detection (20).
Molecular techniques for the diagnosis of D. fragilis show much promise, with PCR demonstrating excellent sensitivity and specificity (18). Such techniques have been used successfully for the diagnosis of other pathogenic protozoa (11, 19).
Molecular genotyping and sequence analysis have demonstrated that D. fragilis exists as two genetically distinct forms (9, 15, 18, 26). Stark et al. (18) sequenced the SSU rRNA gene of seven Australian D. fragilis isolates, and the data generated from the seven showed no variation among them. These observations support the notion that D. fragilis is a clonal species. The sequences from the Australian isolates, however, differed from the sequence of the D. fragilis strain Bi/PA (ATCC 30948; GenBank accession no. U37461) and were found to be similar to those found in a recent study in the Netherlands (15). The true incidence of the wild-type and variant forms in Australia needs to be established and to determine if such variation has any influence on the pathogenicity of the parasite.
A prospective study was undertaken to determine the prevalence and clinical relevance of D. fragilis infections in an Australian population and to determine the genetic diversity of these isolates obtained at the small-subunit (SSU) rRNA gene locus.
MATERIALS AND METHODS
Fecal specimens. All fecal specimens submitted to the Department of Microbiology at St. Vincent's Hospital, Sydney, for investigation of diarrhea from March 2002 until July 2004 were included in the study. One to three specimens per patient were examined. Specimens from outpatients were collected by the patient and submitted to the laboratory as a fresh specimen along with a portion mixed with sodium acetate-acetic acid-formalin (SAF) preservative. Specimens from inpatients or received without a portion fixed in SAF were immediately preserved in SAF upon arrival at the laboratory.
Microbiological investigation. Fecal specimens were cultured for the following bacterial pathogens: Salmonella spp., Shigella spp., Campylobacter spp., Aeromonas spp., Yersinia spp., and Clostridium difficile, and culture for Vibrio spp. was performed where indicated if a patient had history of travel to an endemic area, using standard laboratory procedures and techniques.
An immunochromatographic screening test, the Adeno/Rota STAT-PAK (Chembio Diagnostic Systems Inc., Sydney) for the detection of adenovirus and rotavirus antigen in feces was used according to the manufacturer's recommendations.
Approximately 1 g of feces was placed into SAF and fixed overnight. The fixed specimens were then stained using a modified iron hematoxylin stain, incorporating a carbol fuchsin step to detect coccidia (Fronine, Australia), according to the manufacturer's recommendations. Formalin-ethyl acetate concentration was used for the detection of any helminth ova. In addition, any specimens from human immunodeficiency virus-infected patients were examined for microsporidial spores using the Uvitex 2B stain (21).
PCR for D. fragilis. All specimens where D. fragilis was detected by permanent stain underwent DNA extraction and PCR for D. fragilis-specific DNA using primers DF400 and DF1250 as previously described (18).
Restriction fragment length polymorphism. Restriction fragment length polymorphism (RFLP) analysis was undertaken on all positive PCR products. Eight μl of the PCR product was digested with 10 U of DdeI (Roche, Australia) in a final volume of 15 μl for 1 h at 37°C. Samples were analyzed by electrophoresis on 3% ReadyAgarose gels (Bio-Rad, Sydney).
Follow-up data. Clinical data were collected from all patients diagnosed with D. fragilis. Wherever possible multiple, sticky-tape tests, two to five tapes per patient, were conducted for the detection of Enterobius vermicularis.
Control group. A control group comprising 900 fecal samples from patients without diarrhea or symptoms of gastroenteritis (submitted for occult blood testing and fecal reducing substances) were used. These specimens were processed as above and stained using a modified iron hematoxylin stain. Ninety of these specimens were randomly selected and underwent PCR using D. fragilis-specific primers as described by Stark et al. (18).
Questionnaire. Questionnaires were distributed to 26 laboratories in the Sydney metropolitan area. Information requested for the calendar years 1996 to 2002 included total number of fecal samples processed for ova cysts and parasites, total number of specimens positive for D. fragilis, use of permanent stain, fixation method used in this period, and the situation in which a fixation method would be used.
RESULTS
A total of 6,750 patients submitted fecal specimens between March 2002 and July 2004. Sixty patients were diagnosed with D. fragilis infection from the permanent stains, giving a prevalence of infection of 0. 9%. The results found in this study are summarized in Table 1.
Of the 60 patients infected with D. fragilis, six (10%) had a history of recent overseas travel; three to Southeast Asia, one to Timor, one to Fiji, and one to Papua New Guinea. The remaining 54 of 60 patients (90%) had no recent history of travel outside Australia.
A total of 24/60 (40%) patients had other parasites detected (Table 2). No coccidian parasites were detected. The only other pathogenic protozoan was Giardia intestinalis, which was found concurrently with D. fragilis in three samples. The remaining 36 patients (60%) had only D. fragilis detected. All fecal samples were semiformed or liquid.
The most frequent clinical symptoms associated with D. fragilis infection were diarrhea, abdominal pain, and loose bowel movements. Vomiting was only reported in one patient. Chronic persistent symptoms were common, with 19/60 (32%) patients having diarrhea of over 2 weeks in duration, and one patient claimed to have intermittent diarrhea for several years. Five patients had recurrent D. fragilis infections. One patient was diagnosed with irritable bowel syndrome. All patients were symptomatic. Only one patient (human immunodeficiency virus infected) was immunosuppressed, with all the others being immunocompetent. No microsporidia were detected in the human immunodeficiency virus-infected patient.
Thirty patients were female and 30 were male, with the age range being 3 to 79 years (Fig. 1). The average age was 39.8 years, with a median of 44.5 years. No seasonal variation was found with D. fragilis infection.
No helminth ova were detected in the 60 patients using a formalin-ethyl acetate concentration technique, and no Enterobius vermicularis adults or ova were found; 33/60 (55%) patients submitted a sticky-tape test for E. vermicularis ova, all of which were negative.
No bacterial pathogens were isolated from the patients with D. fragilis infection. The immunochromatographic tests for both adenovirus and rotavirus were also negative for all of the patients.
PCR was performed on 54 of the 60 samples; for six specimens there was a delay (>7 days) in undertaking the DNA extraction, so these specimens were excluded from PCR testing. A specific D. fragilis PCR product of approximately 870 bp was detected in 50 out of 54 samples using the D. fragilis-specific primers designed by Stark et al. (18). RFLP was performed on the 50 positive PCR samples. All gave identical RFLP patterns (data not shown).
Nine hundred fecal samples from patients without gastrointestinal symptoms were used as a control group. No D. fragilis was detected by permanent staining. However, nonpathogenic protozoa were detected in the control group. Blastocystis hominis was found in 47 (5.2%) patients and Endolimax nana in 19 (2.1%), while Blastocystis hominis and Endolimax nana were found concurrently in 12 (1.3%) patients. One patient (0.1%) was found to have Entamoeba hartmanni. PCR using D. fragilis-specific primers was undertaken on 90 samples randomly chosen from the control group. All 90 specimens were negative for D. fragilis DNA by PCR.
Of the 26 laboratories that were sent the questionnaire, only 11 responded. The remaining 15 laboratories were contacted, and four agreed to participate in a phone interview using the same questions as on the written questionnaire. Of the 15 laboratories it was determined that only three in the Sydney metropolitan area routinely performed permanent stains on feces for ova, cysts, and parasite examinations.
DISCUSSION
Dientamoeba fragilis has a worldwide cosmopolitan distribution. In Australia and New Zealand, the reported prevalence rate ranges from 0.4% in western Australia (1) and 1.5% in an urban community in Brisbane (17) to 2.2% in Christchurch, New Zealand (14), and 16.8% in suburban Sydney (21). A longitudinal study of parasite infections in Aboriginal children from the Queensland outback found a prevalence of 5.0% for D. fragilis (23). In this present study a prevalence of 0.9% was found; this is in stark contrast to the prevalence of 16.8% that was found by Walker et al. (22) in the Sydney suburb of French's Forrest.
In this study, D. fragilis infection was closely associated with diarrhea, abdominal pain, and loose bowel movements. All patients with D. fragilis infection were symptomatic, and bacterial and viral causes of these symptoms are unlikely, as routine microbiological cultures and adenovirus and rotavirus testing were negative. However, testing for pathogenic Escherichia coli or norovirus was not undertaken. Three patients were also infected with Giardia intestinalis, which could have caused the gastrointestinal symptoms described in those patients.
One important finding of this study was that chronic persistent symptoms were common. Thirty-two percent of patients had diarrhea for more than 2 weeks, and one patient claimed to have had intermittent diarrhea for several years. Five patients had recurrence of symptoms during the course of the study. It is unknown whether these recurrences were due to treatment failure or reinfection from a common source. One patient was diagnosed with irritable bowel syndrome and was subsequently found to have D. fragilis infection. A recent Australian study by Borody et al. (2) showed a link between D. fragilis and irritable bowel syndrome. Twenty-one patients diagnosed with irritable bowel syndrome and concurrent D. fragilis infection were treated with iodoquinol and doxycycline. Complete elimination of D. fragilis with marked clinical improvement occurred in the majority of patients.
Ten percent of patients diagnosed with D. fragilis infection had a history of recent overseas travel, including Southeast Asia, Papua New Guinea, Timor, and Fiji. Dientamoeba fragilis has been implicated as a cause of diarrhea in returning Swedish travelers, with Norberg et al. (13) finding 63% of patients in a retrospective study had been infected outside the country. Most patients were infected in Africa, South America, and the Middle East.
No parasites were detected by formalin-ethyl acetate concentrations performed on fecal specimens from the D. fragilis-infected patients. Fifty-five percent of the patients submitted multiple tape tests for the detection of Enterobius vermicularis ova, and no E. vermicularis ova were detected. Many researchers have postulated that pinworm is a vector for D. fragilis transmission. Burrows and Swerdlow (3) were the first to propose that E. vermicularis might be a vector for D. fragilis. Several other researchers also found a higher than expected concurrence of D. fragilis and E. vermicularis coinfections (4, 5, 16, 26).
In contrast, a recent study of 25 pediatric cases of D. fragilis found no infections were associated with E. vermicularis (7). These results, along with the findings from this present study, would argue against the hypothesis that E. vermicularis plays a significant role in the transmission of D. fragilis. Most studies that have examined D. fragilis infection have inadequately examined for E. vermicularis. It has yet to be proven what role helminth ova play in the transmission of D. fragilis. Further study is required to ascertain the true mode of transmission of this organism.
Other enteric protozoa were present in 40% of patients with D. fragilis infection. The most common organism was B. hominis. Other protozoa present included E. nana, E. hominis, E. coli, Iodamoeba butschlii, and G. intestinalis. All of these parasites are known to be transmitted via the fecal-oral route. Other researchers have found similar rates of coinfection of D. fragilis with other parasites that are transmitted via the fecal-oral route. Windsor et al. (25) found 54% of patients with D. fragilis had other parasites or enteropathogens present. These findings provide circumstantial evidence to support a hypothesis for a fecal-oral route of transmission for D. fragilis.
No D. fragilis trophozoites were detected in the control group of 900 smaples from patients without gastrointestinal symptoms. This is in contrast to other studies, where D. fragilis was detected in patients with no clinical symptoms (6) and in a case-control study on gastroenteritis from the Netherlands, where D. fragilis was recovered more frequently from controls than case patients (8). These findings may be attributed to the fact that asymptomatic carriage of intestinal protozoa can often occur.
The permanent stained smears positive for D. fragilis were confirmed by PCR. A sensitivity of 93% (50/54 samples) was obtained using a previously published method (18). All 90 negative samples from the control group failed to produce a PCR product.
Sequence data generated in several studies supports the notion for at least two distinct genetic variants of D. fragilis. Johnson and Clarke (9) estimated a sequence divergence of 2% between the two SSU rRNA genotypes of D. fragilis; this was later supported by Peek et al. (15) by sequencing a 558bp region of the SSU rDNA. Sequence data generated by Stark et al. (18) from the entire SSU rDNA region of Australian isolates of D. fragilis showed a greater sequence divergence of 4% between the Australian genotypes and the D. fragilis strain Bi/PA (ATCC 30948). All Australian strains sequenced were identical, which supports the notion that D. fragilis is a clonal species.
The Australian isolates were found to be similar to those found in a recent study in the Netherlands and do not contain the polymorphic DdeI restriction site (CTTAG) at position 644 found in D. fragilis strain Bi/PA (15). RFLP analysis was undertaken on all 50 Australian samples to determine the genotypes present in the Australian population and the extent of genetic diversity. The PCR used in this study amplifies the SSU rRNA region from approximately position 400 to position 1270. This PCR product contains two DdeI restriction sites (CTTAG) that are present in the D. fragilis ATCC 30948 strain yet are absent in the Australian genotypes. All 50 D. fragilis samples showed no variation and corresponded to genotype A. These findings suggest that D. fragilis in Sydney, Australia, displays only a single genotype in fecal samples from various groups, including inpatients, outpatients, and travelers. Further studies are needed to identify the presence of other genotypes throughout Australia.
Dientamoeba fragilis has no recognized cyst stage, and as such, diagnosis is dependent on detecting the trophozoites. As these trophozoites degenerate rapidly prompt fixation of the specimen is necessary (26). Successful diagnosis of D. fragilis is closely associated with the use of permanent stains of fecal smears. Failure to use permanent staining and fixation techniques will inevitably preclude identification of D. fragilis. The aim of the questionnaire sent to the Sydney laboratories was to determine how many laboratories routinely undertake permanent staining and therefore how many laboratories are able to report the presence of D. fragilis. Of the 26 Sydney laboratories, 58% (15/26) participated in the survey, and only three routinely performed permanent staining for ova, cyst and parasites on fecal specimens. Those three laboratories were the only institutions that detected D. fragilis in routine samples. Therefore the true extent of D. fragilis infection must be greatly underestimated as most laboratories do not use techniques to adequately identify this organism.
This is the first prospective study of D. fragilis in Australia to examine clinical data in addition to the genetic diversity of the isolates. Diagnosis was based on permanent staining of fixed fecal smears and confirmed by PCR which demonstrated good sensitivity. All patients infected with D. fragilis were symptomatic and D. fragilis infections were most commonly associated with diarrhea and abdominal pain. Concurrent infections with other protozoa were common, occurring in 40% of samples. The occurrence of D. fragilis with other protozoa that are transmitted via the fecal-oral route would strengthen the case for D. fragilis also being transmitted via this route. No correlation was found with E. vermicularis or any other helminths, questioning the role, if any, pinworm has in the transmission of D. fragilis.
The genetic diversity within 50 samples was examined by PCR followed by RFLP. These data indicated that a single genotype of D. fragilis was represented, one that is genetically different from the North American D. fragilis strain Bi/PA (ATCC 30948). The evidence that D. fragilis is a pathogen is overwhelming (2, 10, 13, 16, 26), and as such all laboratories should attempt to identify this protozoan by the use of permanent staining techniques or molecular methods.
ACKNOWLEDGMENTS
This work was supported by a grant from the Institute of Laboratory Medicine at St. Vincent's Hospital, Sydney, Australia.
REFERENCES
Anonymous.1992 . Western Australian enteric pathogen report.Communicable Dis. Intell. 16:154-159.
Borody, T. J., E. F. Warren, A. Wettstein, G. Robertson, P. Recabarren, A. Fontella, K. Herdnman, and R. Surace.2002 . Eradication of Dientamoeba fragilis can resolve IBS-like symptoms. J. Gastroenterol. Hepatol. 17(Suppl.):A103.
Burrows, R. B., and M. A. Swerdlow. 1956. Enterobius vermicularis as a probable vector of Dientamoeba fragilis. Am. J. Trop. Med. Hyg. 5:258-265.
Burrows, R. B., M. A. Swerdlow, J. K. Frost, and C. K. Leeper. 1954. Pathology of Dientamoeba fragilis infections in the appendix.Am. J. Trop. Med. Hyg. 3:1033-1039.
Cerva, L., M. Schrottenbaum, and V. Kliment. 1991. Intestinal parasites: a study of human appendices. Fiola Parasitol. 38:5-9.
Colea, A., R. Silard, D. Panaitescu, P. Florescu, N. Roman, and T. Capraru. 1980. Studies on Dientamoeba fragilis in Romania. II. Incidence of Dientamoeba in healthy persons. Arch. Roum. Pathol. Exp. Microbiol. 39:49-53.
Cuffari, C., L. Oligny, and E. G. Seidman. 1998. Dientamoeba fragilis masquerading as allergic colitis.J. Paediatr. Gastroenterol. Nutr. 26:16-20.
De Wit, M. A. S., M. P. G. Koopmans, L. M. Kortbeek, N. J. van Leeuween, J. Vinje, and Y. T. H. P. van Duynhoven. 2001. Etiology of gastroenteritis in sentinel general practises in the Netherlands. Clin. Infect. Dis. 33:280-288.
Johnson, J. A., and G. C. Clarke. 2001. Cryptic Diversity in Dientamoeba fragilis. J. Clin. Microbiol. 12:4653-4354.
Johnson, E. H., J. J. Windsor, and G. C. Clark. 2004. Emerging from obscurity: biological, clinical, and diagnostic aspects of Dientamoeba fragilis.Clin. Microbiol. Rev. 17:553-570.
Limor, J. R., A. A. Lal, and L. Xiao.2002 . Detection and differentiation of Cryptosporidium parasites that are pathogenic for humans by real-time PCR. J. Clin. Microbiol. 40:2335-2338.
Mendez, O. C., G. Szmulewicz, C. Menghi, S. Torres, G. Gonzalez, and C. Gatta. 1994. Comparison of intestinal parasite infestation indexes among HIV-positive and -negative populations.Medina (Buenos Aires) 54:307-310.
Norberg, A., C. E. Nord, and B. Evengard. 2003. Dientamoeba fragilis-a protozoal infection which may cause severe bowel distress. Clin. Microbiol. Infect. 9:65-68.
Oxner, R., G. P. Paltridge, B. A. Chapman, H. B. Cook, and P. F. Sheppard. 1987. Dientamoeba fragilis: a bowel pathogen N.Z. Med. J. 100:64-65.
Peek, R., F. R. Reedeker, and T. van Gool. 2004. Direct amplification and genotyping of Dientamoeba fragilis from human stool specimens. J. Clin. Microbiol. 42:631-635.
Priess, U., G. Ockert, S. Broemme, and A. Otto. 1991. On the clinical importance of Dientamoeba fragilis infections in children. J. Hyg. Epidemiol. Microbiol. Immunol. 35:27-34.
Sawangjaroen, N., R. Luke, and P. Prociv. 1993. Diagnosis by faecal culture of Dientamoeba fragilis infections in Australian patients with diarrhoea. Trans. R. Soc. Trop. Med. Hyg. 87:163-165.
Stark, D., N. Beebe, D. Marriott, J. T. Ellis, and J. Harkness.2005 . Detection of Dientamoeba fragilis in fresh stool specimens using PCR. Int. J. Parasitol. 35:57-62.
Troll, H., H. Marti, and N. Weiss. 1997. Simple differential detection of Entamoeba histolytica and Entamoeba dispar in fresh stool specimens by sodium acetate-acetic acid-formalin concentration and PCR. J. Clin. Microbiol. 35:1701-1705.
Van Gool, T., R. Weijts, E. Lommerse, and T. G. Mank.2003 . The triple faeces test: an effective tool for the detection of intestinal parasites in routine clinical practice.Eur. J. Clin. Microbiol. Infect. Dis. 22:284-290.
Van Gool, T., F. Snijders, P. Reiss, J. K. Eeftinck Schattenkerk, M. A. Van den Bergh Weerman, J. F. Bartelsman, J. J. Bruins, E. U. Canning, and J. Dankert.1993 . Diagnosis of intestinal and disseminated microsporidial infections in patients with HIV by a new rapid fluorescence technique. J. Clin. Pathol. 46:694-699.
Walker, J. C., G. Bahr, and A. S. Ehl.1985 . Gastrointestinal parasites in Sydney. Med. J. Aust. 11:143-480.
Welch, J. S., and J. E. Stuart. 1976. A longitudinal study of parasite infections in 120 Queensland Aboriginal children. Med. J. Aust. 44:14-16.
Windsor, J. J., C. G. Clark, and L. Macfarlane.2004 . Molecular typing of Dientamoeba fragilis.Br. J. Biomed. Sci. 61:152-155.
Windsor, J. J., A. M. Rafay, A. K. Shenoy, and E. H. Johnson. 1998. Incidence of Dientamoeba fragilis in faecal samples submitted for routine microbiology analysis. Br. J. Biomed. Sci. 55:172-175.
Yang, J., and T. H. Scholten. 1977. Dientamoeba fragilis: a review with notes on its epidemiology, pathogenicity, mode of transmission and diagnosis. Am. J. Trop. Med. Hyg. 26:16-22.(D. Stark, N. Beebe, D. Ma)